ICPR2020 Paper Browser

Paper download is intended for registered attendees only, and is subjected to the IEEE Copyright Policy. Any other use is strongly forbidden.

Transformer Networks for Trajectory Forecasting

Francesco Giuliari, Hasan Irtiza, Marco Cristani, Fabio Galasso

Auto-TLDR; TransformerNetworks for Trajectory Prediction of People Interactions

Abstract Slides Poster

Most recent successes on forecasting the people mo-tion are based on LSTM models andallmost recent progress hasbeen achieved by modelling the social interaction among peopleand the people interaction with the scene. We question the useof the LSTM models and propose the novel use of TransformerNetworks for trajectory forecasting. This is a fundamental switchfrom the sequential step-by-step processing of LSTMs to theonly-attention-based memory mechanisms of Transformers. Inparticular, we consider both the original Transformer Network(TF) and the larger Bidirectional Transformer (BERT), state-of-the-art on all natural language processing tasks. Our proposedTransformers predict the trajectories of the individual peoplein the scene. These are “simple” models because each personis modelled separately without any complex human-human norscene interaction terms. In particular, the TF modelwithoutbells and whistlesyields the best score on the largest and mostchallenging trajectory forecasting benchmark of TrajNet [1]. Ad-ditionally, its extension which predicts multiple plausible futuretrajectories performs on par with more engineered techniqueson the 5 datasets of ETH [2]+UCY [3]. Finally, we showthat Transformers may deal with missing observations, as itmay be the case with real sensor data. Code is available atgithub.com/FGiuliari/Trajectory-Transformer

Similar papers

AG-GAN: An Attentive Group-Aware GAN for Pedestrian Trajectory Prediction

Yue Song, Niccolò Bisagno, Syed Zohaib Hassan, Nicola Conci

Auto-TLDR; An attentive group-aware GAN for motion prediction in crowded scenarios

Abstract Slides Poster Similar

Understanding human behaviors in crowded scenarios requires analyzing not only the position of the subjects in space, but also the scene context. Existing approaches mostly rely on the motion history of each pedestrian and model the interactions among people by considering the entire surrounding neighborhood. In our approach, we address the problem of motion prediction by applying coherent group clustering and a global attention mechanism on the LSTM-based Generative Adversarial Networks (GANs). The proposed model consists of an attentive group-aware GAN that observes the agents' past motion and predicts future paths, using (i) a group pooling module to model neighborhood interaction, and (ii) an attention module to specifically focus on hidden states. The experimental results demonstrate that our proposal outperforms state-of-the-art models on common benchmark datasets, and is able to generate socially-acceptable trajectories.

DAG-Net: Double Attentive Graph Neural Network for Trajectory Forecasting

Alessio Monti, Alessia Bertugli, Simone Calderara, Rita Cucchiara

Auto-TLDR; Recurrent Generative Model for Multi-modal Human Motion Behaviour in Urban Environments

Abstract Slides Poster Similar

Understanding human motion behaviour is a critical task for several possible applications like self-driving cars or social robots, and in general for all those settings where an autonomous agent has to navigate inside a human-centric environment. This is non-trivial because human motion is inherently multi-modal: given a history of human motion paths, there are many plausible ways by which people could move in the future. Additionally, people activities are often driven by goals, e.g. reaching particular locations or interacting with the environment. We address both the aforementioned aspects by proposing a new recurrent generative model that considers both single agents’ future goals and interactions between different agents. The model exploits a double attention-based graph neural network to collect information about the mutual influences among different agents and integrates it with data about agents’ possible future objectives. Our proposal is general enough to be applied in different scenarios: the model achieves state-of-the-art results in both urban environments and also in sports applications.

What and How? Jointly Forecasting Human Action and Pose

Yanjun Zhu, Yanxia Zhang, Qiong Liu, Andreas Girgensohn

Auto-TLDR; Forecasting Human Actions and Motion Trajectories with Joint Action Classification and Pose Regression

Abstract Slides Poster Similar

Forecasting human actions and motion trajectories addresses the problem of predicting what a person is going to do next and how they will perform it. This is crucial in a wide range of applications such as assisted living and future co-robotic settings. We propose to simultaneously learn actions and action-related human motion dynamics, while existing works perform them independently. In this paper, we present a method to jointly forecast categories of human action and the pose of skeletal joints in the hope that the two tasks can help each other. As a result, our system can predict not only the future actions but also the motion trajectories that will result. To achieve this, we define a task of joint action classification and pose regression. We employ a sequence to sequence encoder-decoder model combined with multi-task learning to forecast future actions and poses progressively before the action happens. Experimental results on two public datasets, IkeaDB and OAD, demonstrate the effectiveness of the proposed method.

Multiple Future Prediction Leveraging Synthetic Trajectories

Lorenzo Berlincioni, Federico Becattini, Lorenzo Seidenari, Alberto Del Bimbo

Auto-TLDR; Synthetic Trajectory Prediction using Markov Chains

Abstract Slides Poster Similar

Trajectory prediction is an important task, especially in autonomous driving. The ability to forecast the position of other moving agents can yield to an effective planning, ensuring safety for the autonomous vehicle as well for the observed entities. In this work we propose a data driven approach based on Markov Chains to generate synthetic trajectories, which are useful for training a multiple future trajectory predictor. The advantages are twofold: on the one hand synthetic samples can be used to augment existing datasets and train more effective predictors; on the other hand, it allows to generate samples with multiple ground truths, corresponding to diverse equally likely outcomes of the observed trajectory. We define a trajectory prediction model and a loss that explicitly address the multimodality of the problem and we show that combining synthetic and real data leads to prediction improvements, obtaining state of the art results.

Explore and Explain: Self-Supervised Navigation and Recounting

Roberto Bigazzi, Federico Landi, Marcella Cornia, Silvia Cascianelli, Lorenzo Baraldi, Rita Cucchiara

Auto-TLDR; Exploring a Photorealistic Environment for Explanation and Navigation

Abstract Slides Similar

Embodied AI has been recently gaining attention as it aims to foster the development of autonomous and intelligent agents. In this paper, we devise a novel embodied setting in which an agent needs to explore a previously unknown environment while recounting what it sees during the path. In this context, the agent needs to navigate the environment driven by an exploration goal, select proper moments for description, and output natural language descriptions of relevant objects and scenes. Our model integrates a novel self-supervised exploration module with penalty, and a fully-attentive captioning model for explanation. Also, we investigate different policies for selecting proper moments for explanation, driven by information coming from both the environment and the navigation. Experiments are conducted on photorealistic environments from the Matterport3D dataset and investigate the navigation and explanation capabilities of the agent as well as the role of their interactions.

AerialMPTNet: Multi-Pedestrian Tracking in Aerial Imagery Using Temporal and Graphical Features

Maximilian Kraus, Seyed Majid Azimi, Emec Ercelik, Reza Bahmanyar, Peter Reinartz, Alois Knoll

Auto-TLDR; AerialMPTNet: A novel approach for multi-pedestrian tracking in geo-referenced aerial imagery by fusing appearance features

Abstract Slides Poster Similar

Multi-pedestrian tracking in aerial imagery has several applications such as large-scale event monitoring, disaster management, search-and-rescue missions, and as input into predictive crowd dynamic models. Due to the challenges such as the large number and the tiny size of the pedestrians (e.g., 4 x 4 pixels) with their similar appearances as well as different scales and atmospheric conditions of the images with their extremely low frame rates (e.g., 2 fps), current state-of-the-art algorithms including the deep learning-based ones are unable to perform well. In this paper, we propose AerialMPTNet, a novel approach for multi-pedestrian tracking in geo-referenced aerial imagery by fusing appearance features from a Siamese Neural Network, movement predictions from a Long Short-Term Memory, and pedestrian interconnections from a GraphCNN. In addition, to address the lack of diverse aerial multi-pedestrian tracking datasets, we introduce the Aerial Multi-Pedestrian Tracking (AerialMPT) dataset consisting of 307 frames and 44,740 pedestrians annotated. To the best of our knowledge, AerialMPT is the largest and most diverse dataset to this date and will be released publicly. We evaluate AerialMPTNet on AerialMPT and KIT AIS, and benchmark with several state-of-the-art tracking methods. Results indicate that AerialMPTNet significantly outperforms other methods on accuracy and time-efficiency.

Knowledge Distillation for Action Anticipation Via Label Smoothing

Guglielmo Camporese, Pasquale Coscia, Antonino Furnari, Giovanni Maria Farinella, Lamberto Ballan

Auto-TLDR; A Multi-Modal Framework for Action Anticipation using Long Short-Term Memory Networks

Abstract Slides Poster Similar

Human capability to anticipate near future from visual observations and non-verbal cues is essential for developing intelligent systems that need to interact with people. Several research areas, such as human-robot interaction (HRI), assisted living or autonomous driving need to foresee future events to avoid crashes or help people. Egocentric scenarios are classic examples where action anticipation is applied due to their numerous applications. Such challenging task demands to capture and model domain's hidden structure to reduce prediction uncertainty. Since multiple actions may equally occur in the future, we treat action anticipation as a multi-label problem with missing labels extending the concept of label smoothing. This idea resembles the knowledge distillation process since useful information is injected into the model during training. We implement a multi-modal framework based on long short-term memory (LSTM) networks to summarize past observations and make predictions at different time steps. We perform extensive experiments on EPIC-Kitchens and EGTEA Gaze+ datasets including more than 2500 and 100 action classes, respectively. The experiments show that label smoothing systematically improves performance of state-of-the-art models for action anticipation.

Future Urban Scenes Generation through Vehicles Synthesis

Alessandro Simoni, Luca Bergamini, Andrea Palazzi, Simone Calderara, Rita Cucchiara

Auto-TLDR; Predicting the Future of an Urban Scene with a Novel View Synthesis Paradigm

Abstract Slides Poster Similar

In this work we propose a deep learning pipeline to predict the visual future appearance of an urban scene. Despite recent advances, generating the entire scene in an end-to-end fashion is still far from being achieved. Instead, here we follow a two stages approach, where interpretable information is included in the loop and each actor is modelled independently. We leverage a per-object novel view synthesis paradigm; i.e. generating a synthetic representation of an object undergoing a geometrical roto-translation in the 3D space. Our model can be easily conditioned with constraints (e.g. input trajectories) provided by state-of-the-art tracking methods or by the user itself. This allows us to generate a set of diverse realistic futures starting from the same input in a multi-modal fashion. We visually and quantitatively show the superiority of this approach over traditional end-to-end scene-generation methods on CityFlow, a challenging real world dataset.

Towards Practical Compressed Video Action Recognition: A Temporal Enhanced Multi-Stream Network

Bing Li, Longteng Kong, Dongming Zhang, Xiuguo Bao, Di Huang, Yunhong Wang

Auto-TLDR; TEMSN: Temporal Enhanced Multi-Stream Network for Compressed Video Action Recognition

Abstract Slides Poster Similar

Current compressed video action recognition methods are mainly based on completely received compressed videos. However, in real transmission, the compressed video packets are usually disorderly received and lost due to network jitters or congestion. It is of great significance to recognize actions in early phases with limited packets, e.g. forecasting the potential risks from videos quickly. In this paper, we proposed a Temporal Enhanced Multi-Stream Network (TEMSN) for practical compressed video action recognition. First, we use three compressed modalities as complementary cues and build a multi-stream network to capture the rich information from compressed video packets. Second, we design a temporal enhanced module based on Encoder-Decoder structure applied on each stream to infer the missing packets, and generate more complete action dynamics. Thanks to the rich modalities and temporal enhancement, our approach is able to better modeling the action with limited compressed packets. Experiments on HMDB-51 and UCF-101 dataset validate its effectiveness and efficiency.

Self-Supervised Joint Encoding of Motion and Appearance for First Person Action Recognition

Mirco Planamente, Andrea Bottino, Barbara Caputo

Auto-TLDR; A Single Stream Architecture for Egocentric Action Recognition from the First-Person Point of View

Abstract Slides Poster Similar

Wearable cameras are becoming more and more popular in several applications, increasing the interest of the research community in developing approaches for recognizing actions from the first-person point of view. An open challenge in egocentric action recognition is that videos lack detailed information about the main actor's pose and thus tend to record only parts of the movement when focusing on manipulation tasks. Thus, the amount of information about the action itself is limited, making crucial the understanding of the manipulated objects and their context. Many previous works addressed this issue with two-stream architectures, where one stream is dedicated to modeling the appearance of objects involved in the action, and another to extracting motion features from optical flow. In this paper, we argue that learning features jointly from these two information channels is beneficial to capture the spatio-temporal correlations between the two better. To this end, we propose a single stream architecture able to do so, thanks to the addition of a self-supervised block that uses a pretext motion prediction task to intertwine motion and appearance knowledge. Experiments on several publicly available databases show the power of our approach.

Image Sequence Based Cyclist Action Recognition Using Multi-Stream 3D Convolution

Stefan Zernetsch, Steven Schreck, Viktor Kress, Konrad Doll, Bernhard Sick

Auto-TLDR; 3D-ConvNet: A Multi-stream 3D Convolutional Neural Network for Detecting Cyclists in Real World Traffic Situations

Abstract Slides Poster Similar

In this article, we present an approach to detect basic movements of cyclists in real world traffic situations based on image sequences, optical flow (OF) sequences, and past positions using a multi-stream 3D convolutional neural network (3D-ConvNet) architecture. To resolve occlusions of cyclists by other traffic participants or road structures, we use a wide angle stereo camera system mounted at a heavily frequented public intersection. We created a large dataset consisting of 1,639 video sequences containing cyclists, recorded in real world traffic, resulting in over 1.1 million samples. Through modeling the cyclists' behavior by a state machine of basic cyclist movements, our approach takes every situation into account and is not limited to certain scenarios. We compare our method to an approach solely based on position sequences. Both methods are evaluated taking into account frame wise and scene wise classification results of basic movements, and detection times of basic movement transitions, where our approach outperforms the position based approach by producing more reliable detections with shorter detection times. Our code and parts of our dataset are made publicly available.

Context Matters: Self-Attention for Sign Language Recognition

Fares Ben Slimane, Mohamed Bouguessa

Auto-TLDR; Attentional Network for Continuous Sign Language Recognition

Abstract Slides Poster Similar

This paper proposes an attentional network for the task of Continuous Sign Language Recognition. The proposed approach exploits co-independent streams of data to model the sign language modalities. These different channels of information can share a complex temporal structure between each other. For that reason, we apply attention to synchronize and help capture entangled dependencies between the different sign language components. Even though Sign Language is multi-channel, handshapes represent the central entities in sign interpretation. Seeing handshapes in their correct context defines the meaning of a sign. Taking that into account, we utilize the attention mechanism to efficiently aggregate the hand features with their appropriate Spatio-temporal context for better sign recognition. We found that by doing so the model is able to identify the essential Sign Language components that revolve around the dominant hand and the face areas. We test our model on the benchmark dataset RWTH-PHOENIX-Weather 2014, yielding competitive results.

Trajectory-User Link with Attention Recurrent Networks

Tao Sun, Yongjun Xu, Fei Wang, Lin Wu, 塘文钱, Zezhi Shao

Auto-TLDR; TULAR: Trajectory-User Link with Attention Recurrent Neural Networks

Abstract Slides Poster Similar

The prevalent adoptions of GPS-enabled devices have witnessed an explosion of various location-based services which produces a huge amount of trajectories monitoring the individuals' movements. In this paper, we tackle Trajectory-User Link (TUL) problem, which identifies humans' movement patterns and links trajectories to the users who generated them. Existing solutions on TUL problem employ recurrent neural networks and variational autoencoder methods, which face the bottlenecks in the case of excessively long trajectories and fragmentary users' movements. However, these are common characteristics of trajectory data in reality, leading to performance degradation of the existing models. In this paper, we propose an end-to-end attention recurrent neural learning framework, called TULAR (Trajectory-User Link with Attention Recurrent Networks), which focus on selected parts of the source trajectories when linking. TULAR introduce the Trajectory Semantic Vector (TSV) via unsupervised location representation learning and recurrent neural networks, by which to reckon the weight of parts of source trajectory. Further, we employ three attention scores for the weight measurements. Experiments are conducted on two real world datasets and compared with several existing methods, and the results show that TULAR yields a new state-of-the-art performance. Source code is public available at GitHub: https://github.com/taos123/TULAR.

RMS-Net: Regression and Masking for Soccer Event Spotting

Matteo Tomei, Lorenzo Baraldi, Simone Calderara, Simone Bronzin, Rita Cucchiara

Auto-TLDR; An Action Spotting Network for Soccer Videos

Abstract Slides Poster Similar

The recently proposed action spotting task consists in finding the exact timestamp in which an event occurs. This task fits particularly well for soccer videos, where events correspond to salient actions strictly defined by soccer rules (a goal occurs when the ball crosses the goal line). In this paper, we devise a lightweight and modular network for action spotting, which can simultaneously predict the event label and its temporal offset using the same underlying features. We enrich our model with two training strategies: the first one for data balancing and uniform sampling, the second for masking ambiguous frames and keeping the most discriminative visual cues. When tested on the SoccerNet dataset and using standard features, our full proposal exceeds the current state of the art by 3 Average-mAP points. Additionally, it reaches a gain of more than 10 Average-mAP points on the test set when fine-tuned in combination with a strong 2D backbone.

Switching Dynamical Systems with Deep Neural Networks

Cesar Ali Ojeda Marin, Kostadin Cvejoski, Bogdan Georgiev, Ramses J. Sanchez

Auto-TLDR; Variational RNN for Switching Dynamics

Abstract Slides Poster Similar

The problem of uncovering different dynamicalregimes is of pivotal importance in time series analysis. Switchingdynamical systems provide a solution for modeling physical phe-nomena whose time series data exhibit different dynamical modes.In this work we propose a novel variational RNN model forswitching dynamics allowing for both non-Markovian and non-linear dynamical behavior between and within dynamic modes.Attention mechanisms are provided to inform the switchingdistribution. We evaluate our model on synthetic and empiricaldatasets of diverse nature and successfully uncover differentdynamical regimes and predict the switching dynamics.

End-To-End Multi-Task Learning of Missing Value Imputation and Forecasting in Time-Series Data

Jinhee Kim, Taesung Kim, Jang-Ho Choi, Jaegul Choo

Auto-TLDR; Time-Series Prediction with Denoising and Imputation of Missing Data

Abstract Slides Poster Similar

Multivariate time-series prediction is a common task, but it often becomes challenging due to missing values involved in data caused by unreliable sensors and other issues. In fact, inaccurate imputation of missing values can degrade the downstream prediction performance, so it may be better not to rely on the estimated values of missing data. Furthermore, observed data may contain noise, so denoising them can be helpful for the main task at hand. In response, we propose a novel approach that can automatically utilize the optimal combination of the observed and the estimated values to generate not only complete, but also noise-reduced data by our own gating mechanism. We evaluate our model on real-world time-series datasets and achieved state-of-the-art performance, demonstrating that our method successfully handle the incomplete datasets. Moreover, we present in-depth studies using a carefully designed, synthetic multivariate time-series dataset to verify the effectiveness of the proposed model. The ablation studies and the experimental analysis of the proposed gating mechanism show that the proposed method works as an effective denoising as well as imputation method for time-series classification tasks.

PIN: A Novel Parallel Interactive Network for Spoken Language Understanding

Peilin Zhou, Zhiqi Huang, Fenglin Liu, Yuexian Zou

Auto-TLDR; Parallel Interactive Network for Spoken Language Understanding

Abstract Slides Poster Similar

Spoken Language Understanding (SLU) is an essential part of the spoken dialogue system, which typically consists of intent detection (ID) and slot filling (SF) tasks. Recently, recurrent neural networks (RNNs) based methods achieved the state-of-the-art for SLU. It is noted that, in the existing RNN-based approaches, ID and SF tasks are often jointly modeled to utilize the correlation information between them. However, we noted that, so far, the efforts to obtain better performance by supporting bidirectional and explicit information exchange between ID and SF are not well studied. In addition, few studies attempt to capture the local context information to enhance the performance of SF. Motivated by these findings, in this paper, Parallel Interactive Network (PIN) is proposed to model the mutual guidance between ID and SF. Specifically, given an utterance, a Gaussian self-attentive encoder is introduced to generate the context-aware feature embedding of the utterance which is able to capture local context information. Taking the feature embedding of the utterance, Slot2Intent module and Intent2Slot module are developed to capture the bidirectional information flow for ID and SF tasks. Finally, a cooperation mechanism is constructed to fuse the information obtained from Slot2Intent and Intent2Slot modules to further reduce the prediction bias. The experiments on two benchmark datasets, i.e., SNIPS and ATIS, demonstrate the effectiveness of our approach, which achieves a competitive result with state-of-the-art models. More encouragingly, by using the feature embedding of the utterance generated by the pre-trained language model BERT, our method achieves the state-of-the-art among all comparison approaches.

A Grid-Based Representation for Human Action Recognition

Soufiane Lamghari, Guillaume-Alexandre Bilodeau, Nicolas Saunier

Auto-TLDR; GRAR: Grid-based Representation for Action Recognition in Videos

Abstract Slides Poster Similar

Human action recognition (HAR) in videos is a fundamental research topic in computer vision. It consists mainly in understanding actions performed by humans based on a sequence of visual observations. In recent years, HAR have witnessed significant progress, especially with the emergence of deep learning models. However, most of existing approaches for action recognition rely on information that is not always relevant for the task, and are limited in the way they fuse temporal information. In this paper, we propose a novel method for human action recognition that encodes efficiently the most discriminative appearance information of an action with explicit attention on representative pose features, into a new compact grid representation. Our GRAR (Grid-based Representation for Action Recognition) method is tested on several benchmark datasets that demonstrate that our model can accurately recognize human actions, despite intra-class appearance variations and occlusion challenges.

A Novel Attention-Based Aggregation Function to Combine Vision and Language

Matteo Stefanini, Marcella Cornia, Lorenzo Baraldi, Rita Cucchiara

Auto-TLDR; Fully-Attentive Reduction for Vision and Language

Abstract Slides Poster Similar

The joint understanding of vision and language has been recently gaining a lot of attention in both the Computer Vision and Natural Language Processing communities, with the emergence of tasks such as image captioning, image-text matching, and visual question answering. As both images and text can be encoded as sets or sequences of elements - like regions and words - proper reduction functions are needed to transform a set of encoded elements into a single response, like a classification or similarity score. In this paper, we propose a novel fully-attentive reduction method for vision and language. Specifically, our approach computes a set of scores for each element of each modality employing a novel variant of cross-attention, and performs a learnable and cross-modal reduction, which can be used for both classification and ranking. We test our approach on image-text matching and visual question answering, building fair comparisons with other reduction choices, on both COCO and VQA 2.0 datasets. Experimentally, we demonstrate that our approach leads to a performance increase on both tasks. Further, we conduct ablation studies to validate the role of each component of the approach.

VTT: Long-Term Visual Tracking with Transformers

Tianling Bian, Yang Hua, Tao Song, Zhengui Xue, Ruhui Ma, Neil Robertson, Haibing Guan

Auto-TLDR; Visual Tracking Transformer with transformers for long-term visual tracking

Abstract Slides Similar

Long-term visual tracking is a challenging problem. State-of-the-art long-term trackers, e.g., GlobalTrack, utilize region proposal networks (RPNs) to generate target proposals. However, the performance of the trackers is affected by occlusions and large scale or ratio variations. To address these issues, in this paper, we are the first to propose a novel architecture with transformers for long-term visual tracking. Specifically, the proposed Visual Tracking Transformer (VTT) utilizes a transformer encoder-decoder architecture for aggregating global information to deal with occlusion and large scale or ratio variation. Furthermore, it also shows better discriminative power against instance-level distractors without the need for extra labeling and hard-sample mining. We conduct extensive experiments on three largest long-term tracking dataset and have achieved state-of-the-art performance.

Attention-Driven Body Pose Encoding for Human Activity Recognition

Bappaditya Debnath, Swagat Kumar, Marry O'Brien, Ardhendu Behera

Auto-TLDR; Attention-based Body Pose Encoding for Human Activity Recognition

Abstract Slides Poster Similar

This article proposes a novel attention-based body pose encoding for human activity recognition. Most of the existing human activity recognition approaches based on 3D pose data often enrich the input data using additional handcrafted representations such as velocity, super normal vectors, pairwise relations, and so on. The enriched data complements the 3D body joint position data and improves the model performance. In this paper, we propose a novel approach that learns enhanced feature representations from a given sequence of 3D body joints. To achieve this, the approach exploits two body pose streams: 1) a spatial stream which encodes the spatial relationship between various body joints at each time point to learn spatial structure involving the spatial distribution of different body joints 2) a temporal stream that learns the temporal variation of individual body joints over the entire sequence duration to present a temporally enhanced representation. Afterwards, these two pose streams are fused with a multi-head attention mechanism. We also capture the contextual information from the RGB video stream using a deep Convolutional Neural Network (CNN) model combined with a multi-head attention and a bidirectional Long Short-Term Memory (LSTM) network. Finally, the RGB video stream is combined with the fused body pose stream to give a novel end-to-end deep model for effective human activity recognition. The proposed model is evaluated on three datasets including the challenging NTU-RGBD dataset and achieves state-of-the-art results.

KoreALBERT: Pretraining a Lite BERT Model for Korean Language Understanding

Hyunjae Lee, Jaewoong Yun, Bongkyu Hwang, Seongho Joe, Seungjai Min, Youngjune Gwon

Auto-TLDR; KoreALBERT: A monolingual ALBERT model for Korean language understanding

Abstract Slides Poster Similar

Abstract—A Lite BERT (ALBERT) has been introduced to scale-up deep bidirectional representation learning for natural languages. Due to the lack of pretrained ALBERT models for Korean language, the best available practice is the multilingual model or resorting back to the any other BERT-based model. In this paper, we develop and pretrain KoreALBERT, a monolingual ALBERT model specifically for Korean language understanding. We introduce a new training objective, namely Word Order Prediction (WOP), and use alongside the existing MLM and SOP criteria to the same architecture and model parameters. Despite having significantly fewer model parameters (thus, quicker to train), our pretrained KoreALBERT outperforms its BERT counterpart on KorQuAD 1.0 benchmark for machine reading comprehension. Consistent with the empirical results in English by Lan et al., KoreALBERT seems to improve downstream task performance involving multi-sentence encoding for Korean language. The pretrained KoreALBERT is publicly available to encourage research and application development for Korean NLP.

Map-Based Temporally Consistent Geolocalization through Learning Motion Trajectories

Bing Zha, Alper Yilmaz

Auto-TLDR; Exploiting Motion Trajectories for Geolocalization of Object on Topological Map using Recurrent Neural Network

Abstract Slides Poster Similar

In this paper, we propose a novel trajectory learning method that exploits motion trajectories on topological map using recurrent neural network for temporally consistent geolocalization of object. Inspired by human's ability to both be aware of distance and direction of self-motion in navigation, our trajectory learning method learns a pattern representation of trajectories encoded as a sequence of distances and turning angles to assist self-localization. We pose the learning process as a conditional sequence prediction problem in which each output locates the object on a traversable edge in a map. Considering the prediction sequence ought to be topologically connected in the graph-structured map, we adopt two different hypotheses generation and elimination strategies to eliminate disconnected sequence prediction. We demonstrate our approach on the KITTI stereo visual odometry dataset which is a city-scale environment. The key benefits of our approach to geolocalization are that 1) we take advantage of powerful sequence modeling ability of recurrent neural network and its robustness to noisy input, 2) only require a map in the form of a graph and 3) simply use an affordable sensor that generates motion trajectory. The experiments show that the motion trajectories can be learned by training an recurrent neural network, and temporally consistent geolocation can be predicted with both of the proposed strategies.

AttendAffectNet: Self-Attention Based Networks for Predicting Affective Responses from Movies

Thi Phuong Thao Ha, Bt Balamurali, Herremans Dorien, Roig Gemma

Auto-TLDR; AttendAffectNet: A Self-Attention Based Network for Emotion Prediction from Movies

Abstract Slides Poster Similar

In this work, we propose different variants of the self-attention based network for emotion prediction from movies, which we call AttendAffectNet. We take both audio and video into account and incorporate the relation among multiple modalities by applying self-attention mechanism in a novel manner into the extracted features for emotion prediction. We compare it to the typically temporal integration of the self-attention based model, which in our case, allows to capture the relation of temporal representations of the movie while considering the sequential dependencies of emotion responses. We demonstrate the effectiveness of our proposed architectures on the extended COGNIMUSE dataset [1], [2] and the MediaEval 2016 Emotional Impact of Movies Task [3], which consist of movies with emotion annotations. Our results show that applying the self-attention mechanism on the different audio-visual features, rather than in the time domain, is more effective for emotion prediction. Our approach is also proven to outperform state-of-the-art models for emotion prediction.

Dual Path Multi-Modal High-Order Features for Textual Content Based Visual Question Answering

Yanan Li, Yuetan Lin, Hongrui Zhao, Donghui Wang

Auto-TLDR; TextVQA: An End-to-End Visual Question Answering Model for Text-Based VQA

Abstract Slides Similar

As a typical cross-modal problem, visual question answering (VQA) has received increasing attention from the communities of computer vision and natural language processing. Reading and reasoning about texts and visual contents in the images is a burgeoning and important research topic in VQA, especially for the visually impaired assistance applications. Given an image, it aims to predict an answer to a provided natural language question closely related to its textual contents. In this paper, we propose a novel end-to-end textual content based VQA model, which grounds question answering both on the visual and textual information. After encoding the image, question and recognized text words, it uses multi-modal factorized high-order modules and the attention mechanism to fuse question-image and question-text features respectively. The complex correlations among different features can be captured efficiently. To ensure the model's extendibility, it embeds candidate answers and recognized texts in a semantic embedding space and adopts semantic embedding based classifier to perform answer prediction. Extensive experiments on the newly proposed benchmark TextVQA demonstrate that the proposed model can achieve promising results.

Tackling Contradiction Detection in German Using Machine Translation and End-To-End Recurrent Neural Networks

Maren Pielka, Rafet Sifa, Lars Patrick Hillebrand, David Biesner, Rajkumar Ramamurthy, Anna Ladi, Christian Bauckhage

Auto-TLDR; Contradiction Detection in Natural Language Inference using Recurrent Neural Networks

Abstract Slides Poster Similar

Natural Language Inference, and specifically Contradiction Detection, is still an unexplored topic with respect to German text. In this paper, we apply Recurrent Neural Network (RNN) methods to learn contradiction-specific sentence embeddings. Our data set for evaluation is a machine-translated version of the Stanford Natural Language Inference (SNLI) corpus. The results are compared to a baseline using unsupervised vectorization techniques, namely tf-idf and Flair, as well as state-of-the art transformer-based (MBERT) methods. We find that the end-to-end models outperform the models trained on unsupervised embeddings, which makes them the better choice in an empirical use case. The RNN methods also perform superior to MBERT on the translated data set.

Emerging Relation Network and Task Embedding for Multi-Task Regression Problems

Schreiber Jens, Bernhard Sick

Auto-TLDR; A Comparative Study of Multi-Task Learning for Non-linear Time Series Problems

Abstract Slides Poster Similar

Multi-Task learning (MTL) provides state-of-the-art results in many applications of computer vision and natural language processing. In contrast to single-task learning (STL), MTL allows for leveraging knowledge between related tasks improving prediction results on all tasks. However, there is a limited number of comparative studies applied to MTL architectures for regression and time series problems taking recent advances of MTL into account. An intriguing, non-linear time-series problem are day ahead forecasts of the expected power generation for renewable power plants. Therefore, the main contribution of this article is a comparative study of the following recent and relevant MTL architectures: Hard-parameter sharing, cross-stitch network, and sluice network (SN). They are compared to a multi-layer peceptron (MLP) model of similar size in an STL setting. As a additional contribution, we provide a simple, yet practical approach to model task specific information through an embedding layer in an MLP, referred to as task embedding. Further, we contribute a new MTL architecture named emerging relation network (ERN), which can be considered as an extension of the SN. For a solar power dataset, the task embedding achieves the best mean improvement with 8.2%. For two wind and one additional solar dataset, the ERN is the best MTL architecture with improvements up to 11.3%.

Learning to Take Directions One Step at a Time

Qiyang Hu, Adrian Wälchli, Tiziano Portenier, Matthias Zwicker, Paolo Favaro

Auto-TLDR; Generating a Sequence of Motion Strokes from a Single Image

Abstract Slides Poster Similar

We present a method to generate a video sequence given a single image. Because items in an image can be animated in arbitrarily many different ways, we introduce as control signal a sequence of motion strokes. Such control signal can be automatically transferred from other videos, e.g., via bounding box tracking. Each motion stroke provides the direction to the moving object in the input image and we aim to train a network to generate an animation following a sequence of such directions. To address this task we design a novel recurrent architecture, which can be trained easily and effectively thanks to an explicit separation of past, future and current states. As we demonstrate in the experiments, our proposed architecture is capable of generating an arbitrary number of frames from a single image and a sequence of motion strokes. Key components of our architecture are an autoencoding constraint to ensure consistency with the past and a generative adversarial scheme to ensure that images look realistic and are temporally smooth. We demonstrate the effectiveness of our approach on the MNIST, KTH, Human3.6M, Push and Weizmann datasets.

Transfer Learning with Graph Neural Networks for Short-Term Highway Traffic Forecasting

Tanwi Mallick, Prasanna Balaprakash, Eric Rask, Jane Macfarlane

Auto-TLDR; Transfer Learning for Highway Traffic Forecasting on Unseen Traffic Networks

Abstract Slides Poster Similar

Large-scale highway traffic forecasting approaches are critical for intelligent transportation systems. Recently, deep-learning-based traffic forecasting methods have emerged as promising approaches for a wide range of traffic forecasting tasks. However, these methods are specific to a given traffic network and consequently, they cannot be used for forecasting traffic on an unseen traffic network. Previous work has identified diffusion convolutional recurrent neural network (DCRNN), as a state-of-the-art method for highway traffic forecasting. It models the complex spatial and temporal dynamics of a highway network using a graph-based diffusion convolution operation within a recurrent neural network. Currently, DCRNN cannot perform transfer learning because it learns location-specific traffic patterns, which cannot be used for unseen regions of a network or new geographic locations. To that end, we develop TL-DCRNN, a new transfer learning approach for DCRNN, where a single model trained on a highway network can be used to forecast traffic on unseen highway networks. Given a traffic network with a large amount of traffic data, our approach consists of partitioning the traffic network into a number of subgraphs and using a new training scheme that utilizes subgraphs for the DCRNN to marginalize the location-specific information, thus learning the traffic as a function of network connectivity and temporal patterns alone. The resulting trained model can be used to forecast traffic on unseen networks. We demonstrate that TL-DCRNN can learn from San Francisco regional traffic data and forecast traffic on the Los Angeles region and vice versa.

Enriching Video Captions with Contextual Text

Philipp Rimle, Pelin Dogan, Markus Gross

Auto-TLDR; Contextualized Video Captioning Using Contextual Text

Abstract Slides Poster Similar

Understanding video content and generating caption with context is an important and challenging task. Unlike prior methods that typically attempt to generate generic video captions without context, our architecture contextualizes captioning by infusing extracted information from relevant text data. We propose an end-to-end sequence-to-sequence model which generates video captions based on visual input, and mines relevant knowledge such as names and locations from contextual text. In contrast to previous approaches, we do not preprocess the text further, and let the model directly learn to attend over it. Guided by the visual input, the model is able to copy words from the contextual text via a pointer-generator network, allowing to produce more specific video captions. We show competitive performance on the News Video Dataset and, through ablation studies, validate the efficacy of contextual video captioning as well as individual design choices in our model architecture.

Constructing Geographic and Long-term Temporal Graph for Traffic Forecasting

Yiwen Sun, Yulu Wang, Kun Fu, Zheng Wang, Changshui Zhang, Jieping Ye

Auto-TLDR; GLT-GCRNN: Geographic and Long-term Temporal Graph Convolutional Recurrent Neural Network for Traffic Forecasting

Abstract Slides Poster Similar

Traffic forecasting influences various intelligent transportation system (ITS) services and is of great significance for user experience as well as urban traffic control. It is challenging due to the fact that the road network contains complex and time-varying spatial-temporal dependencies. Recently, deep learning based methods have achieved promising results by adopting graph convolutional network (GCN) to extract the spatial correlations and recurrent neural network (RNN) to capture the temporal dependencies. However, the existing methods often construct the graph only based on road network connectivity, which limits the interaction between roads. In this work, we propose Geographic and Long-term Temporal Graph Convolutional Recurrent Neural Network (GLT-GCRNN), a novel framework for traffic forecasting that learns the rich interactions between roads sharing similar geographic or long-term temporal patterns. Extensive experiments on a real-world traffic state dataset validate the effectiveness of our method by showing that GLT-GCRNN outperforms the state-of-the-art methods in terms of different metrics.

STaRFlow: A SpatioTemporal Recurrent Cell for Lightweight Multi-Frame Optical Flow Estimation

Pierre Godet, Alexandre Boulch, Aurélien Plyer, Guy Le Besnerais

Auto-TLDR; STaRFlow: A lightweight CNN-based algorithm for optical flow estimation

Abstract Slides Poster Similar

We present a new lightweight CNN-based algorithm for multi-frame optical flow estimation. Our solution introduces a double recurrence over spatial scale and time through repeated use of a generic "STaR" (SpatioTemporal Recurrent) cell. It includes (i) a temporal recurrence based on conveying learned features rather than optical flow estimates; (ii) an occlusion detection process which is coupled with optical flow estimation and therefore uses a very limited number of extra parameters. The resulting STaRFlow algorithm gives state-of-the-art performances on MPI Sintel and Kitti2015 and involves significantly less parameters than all other methods with comparable results.

SDMA: Saliency Driven Mutual Cross Attention for Multi-Variate Time Series

Yash Garg, K. Selcuk Candan

Auto-TLDR; Salient-Driven Mutual Cross Attention for Intelligent Time Series Analytics

Abstract Slides Poster Similar

Integration of rich sensory technologies into critical applications, such as gesture recognition and building energy optimization, has highlighted the importance of intelligent time series analytics. To accommodate this demand, uni-variate approaches have been extended for multi-variate scenarios, but naive extensions have lead to deterioration in model performances due to their limited ability to capture the information recorded in different variates and complex multi-variate time series patterns’ evolution over time. Furthermore, real-world time series are often contaminated with noisy information. In this paper, we note that a time series often carry robust localized temporal events that could help improve model performance by highlighting the relevant information; however, the lack of sufficient data to train for these events make it impossible for neural architectures to identify and make use of these temporal events. We, therefore, argue that a companion process helping identify salient events in the input time series and driving model’s attention to the associated salient sub-sequences can help with learning a high-performing network. Relying on this observation, we propose a novel Saliency-Driven Mutual Cross Attention (SDMA) framework that extracts localized temporal events and generate a saliency series to complement the input time series. We further propose an architecture which accounts for the mutual cross-talk between the input and saliency series branches where input and saliency series attend each other. Experiments show that the proposed mutually-cross attention framework can offer significant boosts in model performance when compared against non-attentioned, conventionally attentioned, and conventionally cross-attentioned models.

Learning Neural Textual Representations for Citation Recommendation

Thanh Binh Kieu, Inigo Jauregi Unanue, Son Bao Pham, Xuan-Hieu Phan, M. Piccardi

Auto-TLDR; Sentence-BERT cascaded with Siamese and triplet networks for citation recommendation

Abstract Slides Poster Similar

With the rapid growth of the scientific literature, manually selecting appropriate citations for a paper is becoming increasingly challenging and time-consuming. While several approaches for automated citation recommendation have been proposed in the recent years, effective document representations for citation recommendation are still elusive to a large extent. For this reason, in this paper we propose a novel approach to citation recommendation which leverages a deep sequential representation of the documents (Sentence-BERT) cascaded with Siamese and triplet networks in a submodular scoring function. To the best of our knowledge, this is the first approach to combine deep representations and submodular selection for a task of citation recommendation. Experiments have been carried out using a popular benchmark dataset -- the ACL Anthology Network corpus -- and evaluated against baselines and a state-of-the-art approach using metrics such as the MRR and F1@k score. The results show that the proposed approach has been able to outperform all the compared approaches in every measured metric.

Exploring Severe Occlusion: Multi-Person 3D Pose Estimation with Gated Convolution

Renshu Gu, Gaoang Wang, Jenq-Neng Hwang

Auto-TLDR; 3D Human Pose Estimation for Multi-Human Videos with Occlusion

Abstract Slides Similar

3D human pose estimation (HPE) is crucial in human behavior analysis, augmented reality/virtual reality (AR/VR) applications, and self-driving industry. Videos that contain multiple potentially occluded people captured from freely moving monocular cameras are very common in real-world scenarios, while 3D HPE for such scenarios is quite challenging, partially because there is a lack of such data with accurate 3D ground truth labels in existing datasets. In this paper, we propose a temporal regression network with a gated convolution module to transform 2D joints to 3D and recover the missing occluded joints in the meantime. A simple yet effective localization approach is further conducted to transform the normalized pose to the global trajectory. To verify the effectiveness of our approach, we also collect a new moving camera multi-human (MMHuman) dataset that includes multiple people with heavy occlusion captured by moving cameras. The 3D ground truth joints are provided by accurate motion capture (MoCap) system. From the experiments on static-camera based Human3.6M data and our own collected moving-camera based data, we show that our proposed method outperforms most state-of-the-art 2D-to-3D pose estimation methods, especially for the scenarios with heavy occlusions.

SAILenv: Learning in Virtual Visual Environments Made Simple

Enrico Meloni, Luca Pasqualini, Matteo Tiezzi, Marco Gori, Stefano Melacci

Auto-TLDR; SAILenv: A Simple and Customized Platform for Visual Recognition in Virtual 3D Environment

Abstract Slides Poster Similar

Recently, researchers in Machine Learning algorithms, Computer Vision scientists, engineers and others, showed a growing interest in 3D simulators as a mean to artificially create experimental settings that are very close to those in the real world. However, most of the existing platforms to interface algorithms with 3D environments are often designed to setup navigation-related experiments, to study physical interactions, or to handle ad-hoc cases that are not thought to be customized, sometimes lacking a strong photorealistic appearance and an easy-to-use software interface. In this paper, we present a novel platform, SAILenv, that is specifically designed to be simple and customizable, and that allows researchers to experiment visual recognition in virtual 3D scenes. A few lines of code are needed to interface every algorithm with the virtual world, and non-3D-graphics experts can easily customize the 3D environment itself, exploiting a collection of photorealistic objects. Our framework yields pixel-level semantic and instance labeling, depth, and, to the best of our knowledge, it is the only one that provides motion-related information directly inherited from the 3D engine. The client-server communication operates at a low level, avoiding the overhead of HTTP-based data exchanges. We perform experiments using a state-of-the-art object detector trained on real-world images, showing that it is able to recognize the photorealistic 3D objects of our environment. The computational burden of the optical flow compares favourably with the estimation performed using modern GPU-based convolutional networks or more classic implementations. We believe that the scientific community will benefit from the easiness and high-quality of our framework to evaluate newly proposed algorithms in their own customized realistic conditions.

Transformer Reasoning Network for Image-Text Matching and Retrieval

Nicola Messina, Fabrizio Falchi, Andrea Esuli, Giuseppe Amato

Auto-TLDR; A Transformer Encoder Reasoning Network for Image-Text Matching in Large-Scale Information Retrieval

Abstract Slides Poster Similar

Image-text matching is an interesting and fascinating task in modern AI research. Despite the evolution of deep-learning-based image and text processing systems, multi-modal matching remains a challenging problem. In this work, we consider the problem of accurate image-text matching for the task of multi-modal large-scale information retrieval. State-of-the-art results in image-text matching are achieved by inter-playing image and text features from the two different processing pipelines, usually using mutual attention mechanisms. However, this invalidates any chance to extract separate visual and textual features needed for later indexing steps in large-scale retrieval systems. In this regard, we introduce the Transformer Encoder Reasoning Network (TERN), an architecture built upon one of the modern relationship-aware self-attentive architectures, the Transformer Encoder (TE). This architecture is able to separately reason on the two different modalities and to enforce a final common abstract concept space by sharing the weights of the deeper transformer layers. Thanks to this design, the implemented network is able to produce compact and very rich visual and textual features available for the successive indexing step. Experiments are conducted on the MS-COCO dataset, and we evaluate the results using a discounted cumulative gain metric with relevance computed exploiting caption similarities, in order to assess possibly non-exact but relevant search results. We demonstrate that on this metric we are able to achieve state-of-the-art results in the image retrieval task. Our code is freely available at https://github.com/mesnico/TERN.

The Color Out of Space: Learning Self-Supervised Representations for Earth Observation Imagery

Stefano Vincenzi, Angelo Porrello, Pietro Buzzega, Marco Cipriano, Pietro Fronte, Roberto Cuccu, Carla Ippoliti, Annamaria Conte, Simone Calderara

Auto-TLDR; Satellite Image Representation Learning for Remote Sensing

Abstract Slides Poster Similar

The recent growth in the number of satellite images fosters the development of effective deep-learning techniques for Remote Sensing (RS). However, their full potential is untapped due to the lack of large annotated datasets. Such a problem is usually countered by fine-tuning a feature extractor that is previously trained on the ImageNet dataset. Unfortunately, the domain of natural images differs from the RS one, which hinders the final performance. In this work, we propose to learn meaningful representations from satellite imagery, leveraging its high-dimensionality spectral bands to reconstruct the visible colors. We conduct experiments on land cover classification (BigEarthNet) and West Nile Virus detection, showing that colorization is a solid pretext task for training a feature extractor. Furthermore, we qualitatively observe that guesses based on natural images and colorization rely on different parts of the input. This paves the way to an ensemble model that eventually outperforms both the above-mentioned techniques.

Visual Oriented Encoder: Integrating Multimodal and Multi-Scale Contexts for Video Captioning

Bang Yang, Yuexian Zou

Auto-TLDR; Visual Oriented Encoder for Video Captioning

Abstract Slides Poster Similar

Video captioning is a challenging task which aims at automatically generating a natural language description of a given video. Recent researches have shown that exploiting the intrinsic multi-modalities of videos significantly promotes captioning performance. However, how to integrate multi-modalities to generate effective semantic representations for video captioning is still an open issue. Some researchers proposed to learn multimodal features in parallel during the encoding stage. The downside of these methods lies in the neglect of the interaction among multi-modalities and their rich contextual information. In this study, inspired by the fact that visual contents are generally more important for comprehending videos, we propose a novel Visual Oriented Encoder (VOE) to integrate multimodal features in an interactive manner. Specifically, VOE is designed as a hierarchical structure, where bottom layers are utilized to extract multi-scale contexts from auxiliary modalities while the top layer is exploited to generate joint representations by considering both visual and contextual information. Following the encoder-decoder framework, we systematically develop a VOE-LSTM model and evaluate it on two mainstream benchmarks: MSVD and MSR-VTT. Experimental results show that the proposed VOE surpasses conventional encoders and our VOE-LSTM model achieves competitive results compared with state-of-the-art approaches.

Compact and Discriminative Multi-Object Tracking with Siamese CNNs

Claire Labit-Bonis, Jérôme Thomas, Frederic Lerasle

Auto-TLDR; Fast, Light-Weight and All-in-One Single Object Tracking for Multi-Target Management

Abstract Slides Poster Similar

Following the tracking-by-detection paradigm, multiple object tracking deals with challenging scenarios, occlusions or even missing detections; the priority is often given to quality measures instead of speed, and a good trade-off between the two is hard to achieve. Based on recent work, we propose a fast, light-weight tracker able to predict targets position and reidentify them at once, when it is usually done with two sequential steps. To do so, we combine a bounding box regressor with a target-oriented appearance learner in a newly designed and unified architecture. This way, our tracker can infer the targets' image pose but also provide us with a confidence level about target identity. Most of the time, it is also common to filter out the detector outputs with a preprocessing step, throwing away precious information about what has been seen in the image. We propose a tracks management strategy able to balance efficiently between detection and tracking outputs and their associated likelihoods. Simply put, we spotlight a full siamese based single object tracker able to predict both position and appearance features at once with a light-weight and all-in-one architecture, within a balanced overall multi-target management strategy. We demonstrate the efficiency and speed of our system w.r.t the literature on the well-known MOT17 challenge benchmark, and bring to the fore qualitative evaluations as well as state-of-the-art quantitative results.

Text Synopsis Generation for Egocentric Videos

Aidean Sharghi, Niels Lobo, Mubarak Shah

Auto-TLDR; Egocentric Video Summarization Using Multi-task Learning for End-to-End Learning

Abstract Slides Similar

Mass utilization of body-worn cameras has led to a huge corpus of available egocentric video. Existing video summarization algorithms can accelerate browsing such videos by selecting (visually) interesting shots from them. Nonetheless, since the system user still has to watch the summary videos, browsing large video databases remain a challenge. Hence, in this work, we propose to generate a textual synopsis, consisting of a few sentences describing the most important events in a long egocentric videos. Users can read the short text to gain insight about the video, and more importantly, efficiently search through the content of a large video database using text queries. Since egocentric videos are long and contain many activities and events, using video-to-text algorithms results in thousands of descriptions, many of which are incorrect. Therefore, we propose a multi-task learning scheme to simultaneously generate descriptions for video segments and summarize the resulting descriptions in an end-to-end fashion. We Input a set of video shots and the network generates a text description for each shot. Next, visual-language content matching unit that is trained with a weakly supervised objective, identifies the correct descriptions. Finally, the last component of our network, called purport network, evaluates the descriptions all together to select the ones containing crucial information. Out of thousands of descriptions generated for the video, a few informative sentences are returned to the user. We validate our framework on the challenging UT Egocentric video dataset, where each video is between 3 to 5 hours long, associated with over 3000 textual descriptions on average. The generated textual summaries, including only 5 percent (or less) of the generated descriptions, are compared to groundtruth summaries in text domain using well-established metrics in natural language processing.

Revisiting Sequence-To-Sequence Video Object Segmentation with Multi-Task Loss and Skip-Memory

Fatemeh Azimi, Benjamin Bischke, Sebastian Palacio, Federico Raue, Jörn Hees, Andreas Dengel

Auto-TLDR; Sequence-to-Sequence Learning for Video Object Segmentation

Abstract Slides Poster Similar

Video Object Segmentation (VOS) is an active research area of the visual domain. One of its fundamental sub-tasks is semi-supervised / one-shot learning: given only the segmentation mask for the first frame, the task is to provide pixel-accurate masks for the object over the rest of the sequence. Despite much progress in the last years, we noticed that many of the existing approaches lose objects in longer sequences, especially when the object is small or briefly occluded. In this work, we build upon a sequence-to-sequence approach that employs an encoder-decoder architecture together with a memory module for exploiting the sequential data. We further improve this approach by proposing a model that manipulates multi-scale spatio-temporal information using memory-equipped skip connections. Furthermore, we incorporate an auxiliary task based on distance classification which greatly enhances the quality of edges in segmentation masks. We compare our approach to the state of the art and show considerable improvement in the contour accuracy metric and the overall segmentation accuracy.

Evaluation of BERT and ALBERT Sentence Embedding Performance on Downstream NLP Tasks

Hyunjin Choi, Judong Kim, Seongho Joe, Youngjune Gwon

Auto-TLDR; Sentence Embedding Models for BERT and ALBERT: A Comparison and Evaluation

Abstract Slides Poster Similar

Contextualized representations from a pre-trained language model are central to achieve a high performance on downstream NLP task. The pre-trained BERT and A Lite BERT (ALBERT) models can be fine-tuned to give state-of-the-art results in sentence-pair regressions such as semantic textual similarity (STS) and natural language inference (NLI). Although BERT-based models yield the [CLS] token vector as a reasonable sentence embedding, the search for an optimal sentence embedding scheme remains an active research area in computational linguistics. This paper explores on sentence embedding models for BERT and ALBERT. In particular, we take a modified BERT network with siamese and triplet network structures called Sentence-BERT (SBERT) and replace BERT with ALBERT to create Sentence-ALBERT (SALBERT). We also experiment with an outer CNN sentence-embedding network for SBERT and SALBERT. We evaluate performances of all sentence-embedding models considered using the STS and NLI datasets. The empirical results indicate that our CNN architecture improves ALBERT models substantially more than BERT models for STS benchmark. Despite significantly fewer model parameters, ALBERT sentence embedding is highly competitive to BERT in downstream NLP evaluations.

Developing Motion Code Embedding for Action Recognition in Videos

Maxat Alibayev, David Andrea Paulius, Yu Sun

Auto-TLDR; Motion Embedding via Motion Codes for Action Recognition

Abstract Slides Poster Similar

We propose a motion embedding strategy via the motion codes that is a vectorized representation of motions based on their salient mechanical attributes. We show that our motion codes can provide robust motion representation. We train a deep neural network model that learns to embed demonstration videos into motion codes. We integrate the extracted features from the motion embedding model into the current state-of-the-art action recognition model. The obtained model achieved higher accuracy than the baseline on a verb classification task from egocentric videos in EPIC-KITCHENS dataset.

Attentive Visual Semantic Specialized Network for Video Captioning

Jesus Perez-Martin, Benjamin Bustos, Jorge Pérez

Auto-TLDR; Adaptive Visual Semantic Specialized Network for Video Captioning

Abstract Slides Poster Similar

As an essential high-level task of video understanding topic, automatically describing a video with natural language has recently gained attention as a fundamental challenge in computer vision. Previous models for video captioning have several limitations, such as the existence of gaps in current semantic representations and the inexpressibility of the generated captions. To deal with these limitations, in this paper, we present a new architecture that we callAttentive Visual Semantic Specialized Network(AVSSN), which is an encoder-decoder model based on our Adaptive Attention Gate and Specialized LSTM layers. This architecture can selectively decide when to use visual or semantic information into the text generation process. The adaptive gate makes the decoder to automatically select the relevant information for providing a better temporal state representation than the existing decoders. Besides, the model is capable of learning to improve the expressiveness of generated captions attending to their length, using a sentence-length-related loss function. We evaluate the effectiveness of the proposed approach on the Microsoft Video Description(MSVD) and the Microsoft Research Video-to-Text (MSR-VTT) datasets, achieving state-of-the-art performance with several popular evaluation metrics: BLEU-4, METEOR, CIDEr, and ROUGE_L.

Modeling Long-Term Interactions to Enhance Action Recognition

Alejandro Cartas, Petia Radeva, Mariella Dimiccoli

Auto-TLDR; A Hierarchical Long Short-Term Memory Network for Action Recognition in Egocentric Videos

Abstract Slides Poster Similar

In this paper, we propose a new approach to understand actions in egocentric videos that exploit the semantics of object interactions at both frame and temporal levels. At the frame level, we use a region-based approach that takes as input a primary region roughly corresponding to the user hands and a set of secondary regions potentially corresponding to the interacting objects and calculates the action score through a CNN formulation. This information is then fed to a Hierarchical Long Short-Term Memory Network (HLSTM) that captures temporal dependencies between actions within and across shots. Ablation studies thoroughly validate the proposed approach, showing in particular that both levels of the HLSTM architecture contribute to performance improvement. Furthermore, quantitative comparisons show that the proposed approach outperforms the state-of-the-art in terms of action recognition on standard benchmarks, without relying on motion information.

Self-Supervised Learning of Dynamic Representations for Static Images

Siyang Song, Enrique Sanchez, Linlin Shen, Michel Valstar

Auto-TLDR; Facial Action Unit Intensity Estimation and Affect Estimation from Still Images with Multiple Temporal Scale

Abstract Slides Poster Similar

Facial actions are spatio-temporal signals by nature, and therefore their modeling is crucially dependent on the availability of temporal information. In this paper, we focus on inferring such temporal dynamics of facial actions when no explicit temporal information is available, i.e. from still images. We present a novel approach to capture multiple scales of such temporal dynamics, with an application to facial Action Unit (AU) intensity estimation and dimensional affect estimation. In particular, 1) we propose a framework that infers a dynamic representation (DR) from a still image, which captures the bi-directional flow of time within a short time-window centered at the input image; 2) we show that we can train our method without the need of explicitly generating target representations, allowing the network to represent dynamics more broadly; and 3) we propose to apply a multiple temporal scale approach that infers DRs for different window lengths (MDR) from a still image. We empirically validate the value of our approach on the task of frame ranking, and show how our proposed MDR attains state of the art results on BP4D for AU intensity estimation and on SEMAINE for dimensional affect estimation, using only still images at test time.

Subspace Clustering for Action Recognition with Covariance Representations and Temporal Pruning

Giancarlo Paoletti, Jacopo Cavazza, Cigdem Beyan, Alessio Del Bue

Auto-TLDR; Unsupervised Learning for Human Action Recognition from Skeletal Data

Abstract Slides Similar

This paper tackles the problem of human action recognition, defined as classifying which action is displayed in a trimmed sequence, from skeletal data. Albeit state-of-the-art approaches designed for this application are all supervised, in this paper we pursue a more challenging direction: Solving the problem with unsupervised learning. To this end, we propose a novel subspace clustering method, which exploits covariance matrix to enhance the action’s discriminability and a timestamp pruning approach that allow us to better handle the temporal dimension of the data. Through a broad experimental validation, we show that our computational pipeline surpasses existing unsupervised approaches but also can result in favorable performances as compared to supervised methods.