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.

Audio-Visual Predictive Coding for Self-Supervised Visual Representation Learning

Mani Kumar Tellamekala, Michel Valstar, Michael Pound, Timo Giesbrecht

Auto-TLDR; AV-PPC: A Multi-task Learning Framework for Learning Semantic Visual Features from Unlabeled Video Data

Abstract Slides Poster

Self-supervised learning has emerged as a candidate approach to learn semantic visual features from unlabeled video data. In self-supervised learning, intrinsic correspondences between data points are used to define a proxy task that forces the model to learn semantic representations. Most existing proxy tasks applied to video data exploit only either intra-modal (e.g. temporal) or cross-modal (e.g. audio-visual) correspondences separately. In theory, jointly learning both these correspondences may result in richer visual features; but, as we show in this work, doing so is non-trivial in practice. To address this problem, we introduce `Audio-Visual Permutative Predictive Coding' (AV-PPC), a multi-task learning framework designed to fully leverage the temporal and cross-modal correspondences as natural supervision signals. In AV-PPC, the model is trained to simultaneously learn multiple intra- and cross-modal predictive coding sub-tasks. By using visual speech recognition (lip-reading) as the downstream evaluation task, we show that our proposed proxy task can learn higher quality visual features than existing proxy tasks. We also show that AV-PPC visual features are highly data-efficient. Without further finetuning, AV-PPC visual encoder achieves 80.30% spoken word classification rate on the LRW dataset, performing on par with directly or fully supervised visual encoders learned from large amounts of labeled data.

Similar papers

Mutual Alignment between Audiovisual Features for End-To-End Audiovisual Speech Recognition

Hong Liu, Yawei Wang, Bing Yang

Auto-TLDR; Mutual Iterative Attention for Audio Visual Speech Recognition

Abstract Slides Poster Similar

Asynchronization issue caused by different types of modalities is one of the major problems in audio visual speech recognition (AVSR) research. However, most AVSR systems merely rely on up sampling of video or down sampling of audio to align audio and visual features, assuming that the feature sequences are aligned frame-by-frame. These pre-processing steps oversimplify the asynchrony relation between acoustic signal and lip motion, lacking ﬂexibility and impairing the performance of the system. Although there are systems modeling the asynchrony between the modalities, sometimes they fail to align speech and video precisely in some even all noise conditions. In this paper, we propose a mutual feature alignment method for AVSR which can make full use of cross modility information to address the asynchronization issue by introducing Mutual Iterative Attention (MIA) mechanism. Our method can automatically learn an alignment in a mutual way by performing mutual attention iteratively between the audio and visual features, relying on the modified encoder structure of Transformer. Experimental results show that our proposed method obtains absolute improvements up to 20.42% over the audio modality alone depending upon the signal-to-noise-ratio (SNR) level. Better recognition performance can also be achieved comparing with the traditional feature concatenation method under both clean and noisy conditions. It is expectable that our proposed mutual feature alignment method can be easily generalized to other multimodal tasks with semantically correlated information.

Robust Audio-Visual Speech Recognition Based on Hybrid Fusion

Hong Liu, Wenhao Li, Bing Yang

Auto-TLDR; Hybrid Fusion Based AVSR with Residual Networks and Bidirectional Gated Recurrent Unit for Robust Speech Recognition in Noise Conditions

Abstract Slides Poster Similar

The fusion of audio and visual modalities is an important stage of audio-visual speech recognition (AVSR), which is generally approached through feature fusion or decision fusion. Feature fusion can exploit the covariations between features from different modalities effectively, whereas decision fusion shows the robustness of capturing an optimal combination of multi-modality. In this work, to take full advantage of the complementarity of the two fusion strategies and address the challenge of inherent ambiguity in noisy environments, we propose a novel hybrid fusion based AVSR method with residual networks and Bidirectional Gated Recurrent Unit (BGRU), which is able to distinguish homophones in both clean and noisy conditions. Specifically, a simple yet effective audio-visual encoder is used to map audio and visual features into a shared latent space to capture more discriminative multi-modal feature and find the internal correlation between spatial-temporal information for different modalities. Furthermore, a decision fusion module is designed to get final predictions in order to robustly utilize the reliability measures of audio-visual information. Finally, we introduce a combined loss, which shows its noise-robustness in learning the joint representation across various modalities. Experimental results on the largest publicly available dataset (LRW) demonstrate the robustness of the proposed method under various noisy conditions.

Audio-Visual Speech Recognition Using a Two-Step Feature Fusion Strategy

Hong Liu, Wanlu Xu, Bing Yang

Auto-TLDR; A Two-Step Feature Fusion Network for Speech Recognition

Abstract Slides Poster Similar

Lip-reading methods and fusion strategy are crucial for audio-visual speech recognition. In recent years, most approaches involve two separate audio and visual streams with early or late fusion strategies. Such a single-stage fusion method may fail to guarantee the integrity and representativeness of fusion information simultaneously. This paper extends a traditional single-stage fusion network to a two-step feature fusion network by adding an audio-visual early feature fusion (AV-EFF) stream to the baseline model. This method can learn the fusion information of different stages, preserving the original features as much as possible and ensuring the independence of different features. Besides, to capture long-range dependencies of video information, a non-local block is added to the feature extraction part of the visual stream (NL-Visual) to obtain the long-term spatio-temporal features. Experimental results on the two largest public datasets in English (LRW) and Mandarin (LRW-1000) demonstrate our method is superior to other state-of-the-art methods.

Temporally Coherent Embeddings for Self-Supervised Video Representation Learning

Joshua Knights, Ben Harwood, Daniel Ward, Anthony Vanderkop, Olivia Mackenzie-Ross, Peyman Moghadam

Auto-TLDR; Temporally Coherent Embeddings for Self-supervised Video Representation Learning

Abstract Slides Poster Similar

This paper presents TCE: Temporally Coherent Embeddings for self-supervised video representation learning. The proposed method exploits inherent structure of unlabeled video data to explicitly enforce temporal coherency in the embedding space, rather than indirectly learning it through ranking or predictive proxy tasks. In the same way that high-level visual information in the world changes smoothly, we believe that nearby frames in learned representations will benefit from demonstrating similar properties. Using this assumption, we train our TCE model to encode videos such that adjacent frames exist close to each other and videos are separated from one another. Using TCE we learn robust representations from large quantities of unlabeled video data. We thoroughly analyse and evaluate our self-supervised learned TCE models on a downstream task of video action recognition using multiple challenging benchmarks (Kinetics400, UCF101, HMDB51). With a simple but effective 2D-CNN backbone and only RGB stream inputs, TCE pre-trained representations outperform all previous self-supervised 2D-CNN and 3D-CNN trained on UCF101. The code and pre-trained models for this paper can be downloaded at: https://github.com/csiro-robotics/TCE

Unsupervised Co-Segmentation for Athlete Movements and Live Commentaries Using Crossmodal Temporal Proximity

Yasunori Ohishi, Yuki Tanaka, Kunio Kashino

Auto-TLDR; A guided attention scheme for audio-visual co-segmentation

Abstract Slides Poster Similar

Audio-visual co-segmentation is a task to extract segments and regions corresponding to specific events on unlabelled audio and video signals. It is particularly important to accomplish it in an unsupervised way, since it is generally very difficult to manually label all the objects and events appearing in audio-visual signals for supervised learning. Here, we propose to take advantage of temporal proximity of corresponding audio and video entities included in the signals. For this purpose, we newly employ a guided attention scheme to this task to efficiently detect and utilize temporal cooccurrences of audio and video information. The experiments using a real TV broadcasting of Sumo wrestling, a sport event, with live commentaries show that our model can automatically extract specific athlete movements and its spoken descriptions in an unsupervised manner.

Improving Mix-And-Separate Training in Audio-Visual Sound Source Separation with an Object Prior

Quan Nguyen, Simone Frintrop, Timo Gerkmann, Mikko Lauri, Julius Richter

Auto-TLDR; Object-Prior: Learning the 1-to-1 correspondence between visual and audio signals by audio- visual sound source methods

Abstract Slides Similar

The performance of an audio-visual sound source separation system is determined by its ability to separate audio sources given images of the sources and the audio mixture. The goal of this study is to investigate the ability to learn the mapping between the sounds and the images of instruments by audio- visual sound source separation methods based on the state-of-the- art PixelPlayer [1]. Theoretical and empirical analyses illustrate that the PixelPlayer is not properly trained to learn the 1-to- 1 correspondence between visual and audio signals during its mix-and-separate training process. Based on the insights from this analysis, a weakly-supervised method called Object-Prior is proposed and evaluated on two audio-visual datasets. The experimental results show that the proposed Object-Prior method outperforms the PixelPlayer and other baselines in the audio- visual sound source separation task. It is also more robust against asynchronized data, where the frame and the audio do not come from the same video, and recognizes musical instruments based on their sound with higher accuracy than the PixelPlayer. This indicates that learning the 1-to-1 correspondence between visual and audio features of an instrument improves the effectiveness of audio-visual sound source separation.

Talking Face Generation Via Learning Semantic and Temporal Synchronous Landmarks

Aihua Zheng, Feixia Zhu, Hao Zhu, Mandi Luo, Ran He

Auto-TLDR; A semantic and temporal synchronous landmark learning method for talking face generation

Abstract Slides Poster Similar

Abstract—Given a speech clip and facial image, the goal of talking face generation is to synthesize a talking face video with accurate mouth synchronization and natural face motion. Recent progress has proven the effectiveness of the landmarks as the intermediate information during talking face generation. However,the large gap between audio and visual modalities makes the prediction of landmarks challenging and limits generation ability. This paper proposes a semantic and temporal synchronous landmark learning method for talking face generation. First, we propose to introduce a word detector to enforce richer semantic information. Then, we propose to preserve the temporal synchronization and consistency between landmarks and audio via the proposed temporal residual loss. Lastly, we employ a U-Net generation network with adaptive reconstruction loss to generate facial images for the predicted landmarks. Experimental results on two benchmark datasets LRW and GRID demonstrate the effectiveness of our model compared to the state-of-the-art methods of talking face generation.

Single-Modal Incremental Terrain Clustering from Self-Supervised Audio-Visual Feature Learning

Reina Ishikawa, Ryo Hachiuma, Akiyoshi Kurobe, Hideo Saito

Auto-TLDR; Multi-modal Variational Autoencoder for Terrain Type Clustering

Abstract Slides Poster Similar

The key to an accurate understanding of terrain is to extract the informative features from the multi-modal data obtained from different devices. Sensors, such as RGB cameras, depth sensors, vibration sensors, and microphones, are used as the multi-modal data. Many studies have explored ways to use them, especially in the robotics field. Some papers have successfully introduced single-modal or multi-modal methods. However, in practice, robots can be faced with extreme conditions; microphones do not work well in the crowded scenes, and an RGB camera cannot capture terrains well in the dark. In this paper, we present a novel framework using the multi-modal variational autoencoder and the Gaussian mixture model clustering algorithm on image data and audio data for terrain type clustering. Our method enables the terrain type clustering even if one of the modalities (either image or audio) is missing at the test-time. We evaluated the clustering accuracy with a conventional multi-modal terrain type clustering method and we conducted ablation studies to show the effectiveness of our approach.

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.

A Neural Lip-Sync Framework for Synthesizing Photorealistic Virtual News Anchors

Ruobing Zheng, Zhou Zhu, Bo Song, Ji Changjiang

Auto-TLDR; Lip-sync: Synthesis of a Virtual News Anchor for Low-Delayed Applications

Abstract Slides Poster Similar

Lip sync has emerged as a promising technique to generate mouth movements from audio signals. However, synthesizing a high-resolution and photorealistic virtual news anchor with current methods is still challenging. The lack of natural appearance, visual consistency, and processing efficiency is the main issue. In this paper, we present a novel lip-sync framework specially designed for producing a virtual news anchor for a target person. A pair of Temporal Convolutional Networks are used to learn the seq-to-seq mapping from audio signals to mouth movements, followed by a neural rendering model that translates the intermediate face representation to the high-quality appearance. This fully-trainable framework avoids several time-consuming steps in traditional graphics-based methods, meeting the requirements of many low-delay applications. Experiments show that our method has advantages over modern neural-based methods in both visual appearance and processing efficiency.

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.

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.

Person Recognition with HGR Maximal Correlation on Multimodal Data

Yihua Liang, Fei Ma, Yang Li, Shao-Lun Huang

Auto-TLDR; A correlation-based multimodal person recognition framework that learns discriminative embeddings of persons by joint learning visual features and audio features

Abstract Slides Poster Similar

Multimodal person recognition is a common task in video analysis and public surveillance, where information from multiple modalities, such as images and audio extracted from videos, are used to jointly determine the identity of a person. Previous person recognition techniques either use only uni-modal data or only consider shared representations between different input modalities, while leaving the extraction of their relationship with identity information to downstream tasks. Furthermore, real-world data often contain noise, which makes recognition more challenging practical situations. In our work, we propose a novel correlation-based multimodal person recognition framework that is relatively simple but can efficaciously learn supervised information in multimodal data fusion and resist noise. Specifically, our framework learns a discriminative embeddings of persons by joint learning visual features and audio features while maximizing HGR maximal correlation among multimodal input and persons' identities. Experiments are done on a subset of Voxceleb2. Compared with state-of-the-art methods, the proposed method demonstrates an improvement of accuracy and robustness to noise.

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.

Audio-Video Detection of the Active Speaker in Meetings

Francisco Madrigal, Frederic Lerasle, Lionel Pibre, Isabelle Ferrané

Auto-TLDR; Active Speaker Detection with Visual and Contextual Information from Meeting Context

Abstract Slides Poster Similar

Meetings are a common activity that provides certain challenges when creating systems that assist them. Such is the case of the Active speaker detection, which can provide useful information for human interaction modeling, or human-robot interaction. Active speaker detection is mostly done using speech, however, certain visual and contextual information can provide additional insights. In this paper we propose an active speaker detection framework that integrates audiovisual features with social information, from the meeting context. Visual cue is processed using a Convolutional Neural Network (CNN) that captures the spatio-temporal relationships. We analyze several CNN architectures with both cues: raw pixels (RGB images) and motion (estimated with optical flow). Contextual reasoning is done with an original methodology, based on the gaze of all participants. We evaluate our proposal with a public \textcolor{black}{benchmark} in state-of-art: AMI corpus. We show how the addition of visual and context information improves the performance of the active speaker detection.

Let's Play Music: Audio-Driven Performance Video Generation

Hao Zhu, Yi Li, Feixia Zhu, Aihua Zheng, Ran He

Auto-TLDR; APVG: Audio-driven Performance Video Generation Using Structured Temporal UNet

Abstract Slides Poster Similar

We propose a new task named Audio-driven Performance Video Generation (APVG), which aims to synthesize the video of a person playing a certain instrument guided by a given music audio clip. It is a challenging task to generate the high-dimensional temporal consistent videos from low-dimensional audio modality. In this paper, we propose a multi-staged framework to achieve this new task to generate realistic and synchronized performance video from given music. Firstly, we provide both global appearance and local spatial information by generating the coarse videos and keypoints of body and hands from a given music respectively. Then, we propose to transform the generated keypoints to heatmap via a differentiable space transformer, since the heatmap offers more spatial information but is harder to generate directly from audio. Finally, we propose a Structured Temporal UNet (STU) to extract both intra-frame structured information and inter-frame temporal consistency. They are obtained via graph-based structure module, and CNN-GRU based high-level temporal module respectively for final video generation. Comprehensive experiments validate the effectiveness of our proposed framework.

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.

Feature-Supervised Action Modality Transfer

Fida Mohammad Thoker, Cees Snoek

Auto-TLDR; Cross-Modal Action Recognition and Detection in Non-RGB Video Modalities by Learning from Large-Scale Labeled RGB Data

Abstract Slides Poster Similar

This paper strives for action recognition and detection in video modalities like RGB, depth maps or 3D-skeleton sequences when only limited modality-specific labeled examples are available. For the RGB, and derived optical-flow, modality many large-scale labeled datasets have been made available. They have become the de facto pre-training choice when recognizing or detecting new actions from RGB datasets that have limited amounts of labeled examples available. Unfortunately, large-scale labeled action datasets for other modalities are unavailable for pre-training. In this paper, our goal is to recognize actions from limited examples in non-RGB video modalities, by learning from large-scale labeled RGB data. To this end, we propose a two-step training process: (i) we extract action representation knowledge from an RGB-trained teacher network and adapt it to a non-RGB student network. (ii) we then fine-tune the transfer model with available labeled examples of the target modality. For the knowledge transfer we introduce feature-supervision strategies, which rely on unlabeled pairs of two modalities (the RGB and the target modality) to transfer feature level representations from the teacher to the the student network. Ablations and generalizations with two RGB source datasets and two non-RGB target datasets demonstrate that an optical-flow teacher provides better action transfer features than RGB for both depth maps and 3D-skeletons, even when evaluated on a different target domain, or for a different task. Compared to alternative cross-modal action transfer methods we show a good improvement in performance especially when labeled non-RGB examples to learn from are scarce.

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.

Cascade Attention Guided Residue Learning GAN for Cross-Modal Translation

Bin Duan, Wei Wang, Hao Tang, Hugo Latapie, Yan Yan

Auto-TLDR; Cascade Attention-Guided Residue GAN for Cross-modal Audio-Visual Learning

Abstract Slides Poster Similar

Since we were babies, we intuitively develop the ability to correlate the input from different cognitive sensors such as vision, audio, and text. However, in machine learning, this cross-modal learning is a nontrivial task because different modalities have no homogeneous properties. Previous works discover that there should be bridges among different modalities. From neurology and psychology perspective, humans have the capacity to link one modality with another one, e.g., associating a picture of a bird with the only hearing of its singing and vice versa. Is it possible for machine learning algorithms to recover the scene given the audio signal? In this paper, we propose a novel Cascade Attention-Guided Residue GAN (CAR-GAN), aiming at reconstructing the scenes given the corresponding audio signals. Particularly, we present a residue module to mitigate the gap between different modalities progressively. Moreover, a cascade attention guided network with a novel classification loss function is designed to tackle the cross-modal learning task. Our model keeps consistency in the high-level semantic label domain and is able to balance two different modalities. The experimental results demonstrate that our model achieves the state-of-the-art cross-modal audio-visual generation on the challenging Sub-URMP dataset.

Image Representation Learning by Transformation Regression

Xifeng Guo, Jiyuan Liu, Sihang Zhou, En Zhu, Shihao Dong

Auto-TLDR; Self-supervised Image Representation Learning using Continuous Parameter Prediction

Abstract Slides Poster Similar

Self-supervised learning is a thriving research direction since it can relieve the burden of human labeling for machine learning by seeking for supervision from data instead of human annotation. Although demonstrating promising performance in various applications, we observe that the existing methods usually model the auxiliary learning tasks as classification tasks with finite discrete labels, leading to insufficient supervisory signals, which in turn restricts the representation quality. In this paper, to solve the above problem and make full use of the supervision from data, we design a regression model to predict the continuous parameters of a group of transformations, i.e., image rotation, translation, and scaling. Surprisingly, this naive modification stimulates tremendous potential from data and the resulting supervisory signal has largely improved the performance of image representation learning. Extensive experiments on four image datasets, including CIFAR10, CIFAR100, STL10, and SVHN, indicate that our proposed algorithm outperforms the state-of-the-art unsupervised learning methods by a large margin in terms of classification accuracy. Crucially, we find that with our proposed training mechanism as an initialization, the performance of the existing state-of-the-art classification deep architectures can be preferably improved.

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.

Learning Visual Voice Activity Detection with an Automatically Annotated Dataset

Stéphane Lathuiliere, Pablo Mesejo, Radu Horaud

Auto-TLDR; Deep Visual Voice Activity Detection with Optical Flow

Abstract Slides Similar

Visual voice activity detection (V-VAD) uses visual features to predict whether a person is speaking or not. V-VAD is useful whenever audio VAD (A-VAD) is inefficient either because the acoustic signal is difficult to analyze or is simply missing. We propose two deep architectures for V-VAD, one based on facial landmarks and one based on optical flow. Moreover, available datasets, used for learning and for testing V-VAD, lack content variability. We introduce a novel methodology to automatically create and annotate very large datasets in-the-wild, based on combining A-VAD and face detection. A thorough empirical evaluation shows the advantage of training the proposed deep V-VAD models with such a dataset.

Self-Supervised Learning for Astronomical Image Classification

Ana Martinazzo, Mateus Espadoto, Nina S. T. Hirata

Auto-TLDR; Unlabeled Astronomical Images for Deep Neural Network Pre-training

Abstract Slides Poster Similar

In Astronomy, a huge amount of image data is generated daily by photometric surveys, which scan the sky to collect data from stars, galaxies and other celestial objects. In this paper, we propose a technique to leverage unlabeled astronomical images to pre-train deep convolutional neural networks, in order to learn a domain-specific feature extractor which improves the results of machine learning techniques in setups with small amounts of labeled data available. We show that our technique produces results which are in many cases better than using ImageNet pre-training.

Joint Supervised and Self-Supervised Learning for 3D Real World Challenges

Antonio Alliegro, Davide Boscaini, Tatiana Tommasi

Auto-TLDR; Self-supervision for 3D Shape Classification and Segmentation in Point Clouds

Abstract Slides Similar

Point cloud processing and 3D shape understanding are very challenging tasks for which deep learning techniques have demonstrated great potentials. Still further progresses are essential to allow artificial intelligent agents to interact with the real world. In many practical conditions the amount of annotated data may be limited and integrating new sources of knowledge becomes crucial to support autonomous learning. Here we consider several scenarios involving synthetic and real world point clouds where supervised learning fails due to data scarcity and large domain gaps. We propose to enrich standard feature representations by leveraging self-supervision through a multi-task model that can solve a 3D puzzle while learning the main task of shape classification or part segmentation. An extensive analysis investigating few-shot, transfer learning and cross-domain settings shows the effectiveness of our approach with state-of-the-art results for 3D shape classification and part segmentation.

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.

Spatial Bias in Vision-Based Voice Activity Detection

Kalin Stefanov, Mohammad Adiban, Giampiero Salvi

Auto-TLDR; Spatial Bias in Vision-based Voice Activity Detection in Multiparty Human-Human Interactions

Abstract Poster Similar

We present models for automatic vision-based voice activity detection (VAD) in multiparty human-human interactions that are aimed at complementing the acoustic VAD methods. We provide evidence that this type of vision-based VAD models are susceptible to spatial bias in the datasets. The physical settings of the interaction, usually constant throughout data acquisition, determines the distribution of head poses of the participants. Our results show that when the head pose distributions are significantly different in the training and test sets, the performance of the models drops significantly. This suggests that previously reported results on datasets with a fixed physical configuration may overestimate the generalization capabilities of this type of models. We also propose a number of possible remedies to the spatial bias, including data augmentation, input masking and dynamic features, and provide an in-depth analysis of the visual cues used by our models.

Hierarchical Multimodal Attention for Deep Video Summarization

Melissa Sanabria, Frederic Precioso, Thomas Menguy

Auto-TLDR; Automatic Summarization of Professional Soccer Matches Using Event-Stream Data and Multi- Instance Learning

Abstract Slides Poster Similar

The way people consume sports on TV has drastically evolved in the last years, particularly under the combined effects of the legalization of sport betting and the huge increase of sport analytics. Several companies are nowadays sending observers in the stadiums to collect live data of all the events happening on the field during the match. Those data contain meaningful information providing a very detailed description of all the actions occurring during the match to feed the coaches and staff, the fans, the viewers, and the gamblers. Exploiting all these data, sport broadcasters want to generate extra content such as match highlights, match summaries, players and teams analytics, etc., to appeal subscribers. This paper explores the problem of summarizing professional soccer matches as automatically as possible using both the aforementioned event-stream data collected from the field and the content broadcasted on TV. We have designed an architecture, introducing first (1) a Multiple Instance Learning method that takes into account the sequential dependency among events and then (2) a hierarchical multimodal attention layer that grasps the importance of each event in an action. We evaluate our approach on matches from two professional European soccer leagues, showing its capability to identify the best actions for automatic summarization by comparing with real summaries made by human operators.

Multi-Modal Deep Clustering: Unsupervised Partitioning of Images

Guy Shiran, Daphna Weinshall

Auto-TLDR; Multi-Modal Deep Clustering for Unlabeled Images

Abstract Slides Poster Similar

The clustering of unlabeled raw images is a daunting task, which has recently been approached with some success by deep learning methods. Here we propose an unsupervised clustering framework, which learns a deep neural network in an end-to-end fashion, providing direct cluster assignments of images without additional processing. Multi-Modal Deep Clustering (MMDC), trains a deep network to align its image embeddings with target points sampled from a Gaussian Mixture Model distribution. The cluster assignments are then determined by mixture component association of image embeddings. Simultaneously, the same deep network is trained to solve an additional self-supervised task. This pushes the network to learn more meaningful image representations and stabilizes the training. Experimental results show that MMDC achieves or exceeds state-of-the-art performance on four challenging benchmarks. On natural image datasets we improve on previous results with significant margins of up to 11% absolute accuracy points, yielding an accuracy of 70% on CIFAR-10 and 61% on STL-10.

Unsupervised Sound Source Localization From Audio-Image Pairs Using Input Gradient Map

Tomohiro Tanaka, Takahiro Shinozaki

Auto-TLDR; Unsupervised Sound Localization Using Gradient Method

Abstract Slides Poster Similar

Humans easily and routinely identify an image region that corresponds to an observed sound in their daily lives. The task is formulated as an unsupervised sound source localization without using tagged data. Recently, several methods have been proposed that utilize the activation of hidden or output layers of neural networks, such as an attention layer or feature maps in a convolutional neural network (CNN). We propose another strategy that obtains a localization map at the input side, applying the widely used input gradient method. It is computationally efficient and can be easily applied to any existing techniques because it is free from the network structure. Taking advantage of it, we propose a combination method with existing methods for higher sound localization performance. Experiments are performed using the Flickr-SoundNet data set. When a pre-trained image front-end was used, the proposed method gives better results than the attention-based method. For a completely unsupervised condition, the gradient method provides comparable performance as the conventional methods; the best results are obtained by this combination method.

Generative Latent Implicit Conditional Optimization When Learning from Small Sample

Idan Azuri, Daphna Weinshall

Auto-TLDR; GLICO: Generative Latent Implicit Conditional Optimization for Small Sample Learning

Abstract Slides Poster Similar

We revisit the long-standing problem of learning from small sample. The generation of new samples from a small training set of labeled points has attracted increased attention in recent years. In this paper, we propose a novel such method called GLICO (Generative Latent Implicit Conditional Optimization). GLICO learns a mapping from the training examples to a latent space and a generator that generates images from vectors in the latent space. Unlike most recent work, which rely on access to large amounts of unlabeled data, GLICO does not require access to any additional data other than the small set of labeled points. In fact, GLICO learns to synthesize completely new samples for every class using as little as 5 or 10 examples per class, with as few as 10 such classes and no data from unknown classes. GLICO is then used to augment the small training set while training a classifier on the small sample. To this end, our proposed method samples the learned latent space using spherical interpolation (slerp) and generates new examples using the trained generator. Empirical results show that the new sampled set is diverse enough, leading to improvement in image classification in comparison with the state of the art when trained on small samples obtained from CIFAR-10, CIFAR-100, and CUB-200.

End-To-End Triplet Loss Based Emotion Embedding System for Speech Emotion Recognition

Puneet Kumar, Sidharth Jain, Balasubramanian Raman, Partha Pratim Roy, Masakazu Iwamura

Auto-TLDR; End-to-End Neural Embedding System for Speech Emotion Recognition

Abstract Slides Poster Similar

In this paper, an end-to-end neural embedding system based on triplet loss and residual learning has been proposed for speech emotion recognition. The proposed system learns the embeddings from the emotional information of the speech utterances. The learned embeddings are used to recognize the emotions portrayed by given speech samples of various lengths. The proposed system implements Residual Neural Network architecture. It is trained using softmax pre-training and triplet loss function. The weights between the fully connected and embedding layers of the trained network are used to calculate the embedding values. The embedding representations of various emotions are mapped onto a hyperplane, and the angles among them are computed using the cosine similarity. These angles are utilized to classify a new speech sample into its appropriate emotion class. The proposed system has demonstrated 91.67\% and 64.44\% accuracy while recognizing emotions for RAVDESS and IEMOCAP dataset, respectively.

Two-Stream Temporal Convolutional Network for Dynamic Facial Attractiveness Prediction

Nina Weng, Jiahao Wang, Annan Li, Yunhong Wang

Auto-TLDR; 2S-TCN: A Two-Stream Temporal Convolutional Network for Dynamic Facial Attractiveness Prediction

Abstract Slides Poster Similar

In the field of facial attractiveness prediction, while deep models using static pictures have shown promising results, little attention is paid to dynamic facial information, which is proven to be influential by psychological studies. Meanwhile, the increasing popularity of short video apps creates an enormous demand of facial attractiveness prediction from short video clips. In this paper, we target on the dynamic facial attractiveness prediction problem. To begin with, a large-scale video-based facial attractiveness prediction dataset (VFAP) with more than one thousand clips from TikTok is collected. A two-stream temporal convolutional network (2S-TCN) is then proposed to capture dynamic attractiveness feature from both facial appearance and landmarks. We employ attentive feature enhancement along with specially designed modality and temporal fusion strategies to better explore the temporal dynamics. Extensive experiments on the proposed VFAP dataset demonstrate that 2S-TCN has a distinct advantage over the state-of-the-art static prediction methods.

Mutual Information Based Method for Unsupervised Disentanglement of Video Representation

Aditya Sreekar P, Ujjwal Tiwari, Anoop Namboodiri

Auto-TLDR; MIPAE: Mutual Information Predictive Auto-Encoder for Video Prediction

Abstract Slides Poster Similar

Video Prediction is an interesting and challenging task of predicting future frames from a given set context frames that belong to a video sequence. Video prediction models have found prospective applications in Maneuver Planning, Health care, Autonomous Navigation and Simulation. One of the major challenges in future frame generation is due to the high dimensional nature of visual data. In this work, we propose Mutual Information Predictive Auto-Encoder (MIPAE) framework, that reduces the task of predicting high dimensional video frames by factorising video representations into content and low dimensional pose latent variables that are easy to predict. A standard LSTM network is used to predict these low dimensional pose representations. Content and the predicted pose representations are decoded to generate future frames. Our approach leverages the temporal structure of the latent generative factors of a video and a novel mutual information loss to learn disentangled video representations. We also propose a metric based on mutual information gap (MIG) to quantitatively access the effectiveness of disentanglement on DSprites and MPI3D-real datasets. MIG scores corroborate with the visual superiority of frames predicted by MIPAE. We also compare our method quantitatively on evaluation metrics LPIPS, SSIM and PSNR.

Reducing the Variance of Variational Estimates of Mutual Information by Limiting the Critic's Hypothesis Space to RKHS

Aditya Sreekar P, Ujjwal Tiwari, Anoop Namboodiri

Auto-TLDR; Mutual Information Estimation from Variational Lower Bounds Using a Critic's Hypothesis Space

Abstract Slides Similar

Mutual information (MI) is an information-theoretic measure of dependency between two random variables. Several methods to estimate MI, from samples of two random variables with unknown underlying probability distributions have been proposed in the literature. Recent methods realize parametric probability distributions or critic as a neural network to approximate unknown density ratios. The approximated density ratios are used to estimate different variational lower bounds of MI. While these methods provide reliable estimation when the true MI is low, they produce high variance estimates in cases of high MI. We argue that the high variance characteristic is due to the uncontrolled complexity of the critic's hypothesis space. In support of this argument, we use the data-driven Rademacher complexity of the hypothesis space associated with the critic's architecture to analyse generalization error bound of variational lower bound estimates of MI. In the proposed work, we show that it is possible to negate the high variance characteristics of these estimators by constraining the critic's hypothesis space to Reproducing Hilbert Kernel Space (RKHS), which corresponds to a kernel learned using Automated Spectral Kernel Learning (ASKL). By analysing the aforementioned generalization error bounds, we augment the overall optimisation objective with effective regularisation term. We empirically demonstrate the efficacy of this regularization in enforcing proper bias variance tradeoff on four variational lower bounds, namely NWJ, MINE, JS and SMILE.

Video Semantic Segmentation Using Deep Multi-View Representation Learning

Akrem Sellami, Salvatore Tabbone

Auto-TLDR; Deep Multi-view Representation Learning for Video Object Segmentation

Abstract Slides Poster Similar

In this paper, we propose a deep learning model based on deep multi-view representation learning, to address the video object segmentation task. The proposed model emphasizes the importance of the inherent correlation between video frames and incorporates a multi-view representation learning based on deep canonically correlated autoencoders. The multi-view representation learning in our model provides an efficient mechanism for capturing inherent correlations by jointly extracting useful features and learning better representation into a joint feature space, i.e., shared representation. To increase the training data and the learning capacity, we train the proposed model with pairs of video frames, i.e., $F_{a}$ and $F_{b}$. During the segmentation phase, the deep canonically correlated autoencoders model encodes useful features by processing multiple reference frames together, which is used to detect the frequently reappearing. Our model enhances the state-of-the-art deep learning-based methods that mainly focus on learning discriminative foreground representations over appearance and motion. Experimental results over two large benchmarks demonstrate the ability of the proposed method to outperform competitive approaches and to reach good performances, in terms of semantic segmentation.

Learning Group Activities from Skeletons without Individual Action Labels

Fabio Zappardino, Tiberio Uricchio, Lorenzo Seidenari, Alberto Del Bimbo

Auto-TLDR; Lean Pose Only for Group Activity Recognition

Abstract Similar

To understand human behavior we must not just recognize individual actions but model possibly complex group activity and interactions. Hierarchical models obtain the best results in group activity recognition but require fine grained individual action annotations at the actor level. In this paper we show that using only skeletal data we can train a state-of-the art end-to-end system using only group activity labels at the sequence level. Our experiments show that models trained without individual action supervision perform poorly. On the other hand we show that pseudo-labels can be computed from any pre-trained feature extractor with comparable final performance. Finally our carefully designed lean pose only architecture shows highly competitive results versus more complex multimodal approaches even in the self-supervised variant.

Three-Dimensional Lip Motion Network for Text-Independent Speaker Recognition

Jianrong Wang, Tong Wu, Shanyu Wang, Mei Yu, Qiang Fang, Ju Zhang, Li Liu

Auto-TLDR; Lip Motion Network for Text-Independent and Text-Dependent Speaker Recognition

Abstract Slides Poster Similar

Lip motion reflects behavior characteristics of speakers, and thus can be used as a new kind of biometrics in speaker recognition. In the literature, lots of works used two dimensional (2D) lip images to recognize speaker in a text-dependent context. However, 2D lip easily suffers from face orientations. To this end, in this work, we present a novel end-to-end 3D lip motion Network (3LMNet) by utilizing the sentence-level 3D lip motion (S3DLM) to recognize speakers in both the text-independent and text-dependent contexts. A novel regional feedback module (RFM) is proposed to explore attentions in different lip regions. Besides, prior knowledge of lip motion is investigated to complement RFM, where landmark-level and frame-level features are merged to form a better feature representation. Moreover, we present two methods, i.e., coordinate transformation and face posture correction to pre-process the LSD-AV dataset, which contains 68 speakers and 146 sentences per speaker. The evaluation results on this dataset demonstrate that our proposed 3LMNet is superior to the baseline models, i.e., LSTM, VGG-16 and ResNet-34, and outperforms the state-of-the-art using 2D lip image as well as the 3D face. The code of this work is released at https://github.com/wutong18/Three-Dimensional-Lip-Motion-Ne twork-for-Text-Independent-Speaker-Recognition.

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.

Not 3D Re-ID: Simple Single Stream 2D Convolution for Robust Video Re-Identification

Toby Breckon, Aishah Alsehaim

Auto-TLDR; ResNet50-IBN for Video-based Person Re-Identification using Single Stream 2D Convolution Network

Abstract Slides Poster Similar

Video-based person re-identification has received increasing attention recently, as it plays an important role within the surveillance video analysis. Video-based Re-ID is an expansion of earlier image-based re-identification methods by learning features from a video via multiple image frames for each person. Most contemporary video Re-ID methods utilise complex CNN-based network architectures using 3D convolution or multi-branch networks to extract spatial-temporal features from the video. By contrast, in this paper, we will illustrate superior performance from a simple single stream 2D convolution network leveraging the ResNet50-IBN architecture to extract frame-level features followed by temporal attention for clip level features. These clip level features can be generalised to extract video level features by averaging clip level features without any additional cost. Our model, uses best video Re-ID practice and transfer learning between datasets, outperforms existing state-of-the-art approaches on MARS, PRID2011 and iLIDSVID datasets with 89:62%, 97:75%, 97:33% rank-1 accuracy respectively and with 84:61% mAP for MARS, without reliance on complex and memory intensive 3D convolutions or multistream networks architectures as found in other contemporary work. Conversely, this work shows that global features extracted by the 2D convolution network are a sufficient representation for robust state of the art video Re-ID.

Multi-Scale 2D Representation Learning for Weakly-Supervised Moment Retrieval

Ding Li, Rui Wu, Zhizhong Zhang, Yongqiang Tang, Wensheng Zhang

Auto-TLDR; Multi-scale 2D Representation Learning for Weakly Supervised Video Moment Retrieval

Abstract Slides Poster Similar

Video moment retrieval aims to search the moment most relevant to a given language query. However, most existing methods in this community often require temporal boundary annotations which are expensive and time-consuming to label. Hence weakly supervised methods have been put forward recently by only using coarse video-level label. Despite effectiveness, these methods usually process moment candidates independently, while ignoring a critical issue that the natural temporal dependencies between candidates in different temporal scales. To cope with this issue, we propose a Multi-scale 2D Representation Learning method for weakly supervised video moment retrieval. Specifically, we first construct a two-dimensional map for each temporal scale to capture the temporal dependencies between candidates. Two dimensions in this map indicate the start and end time points of these candidates. Then, we select top-K candidates from each scale-varied map with a learnable convolutional neural network. With a newly designed Moments Evaluation Module, we obtain the alignment scores of the selected candidates. At last, the similarity between captions and language query is served as supervision for further training the candidates' selector. Experiments on two benchmark datasets Charades-STA and ActivityNet Captions demonstrate that our approach achieves superior performance to state-of-the-art results.

Semi-Supervised Class Incremental Learning

Alexis Lechat, Stéphane Herbin, Frederic Jurie

Auto-TLDR; incremental class learning with non-annotated batches

Abstract Slides Poster Similar

This paper makes a contribution to the problem of incremental class learning, the principle of which is to sequentially introduce batches of samples annotated with new classes during the learning phase. The main objective is to reduce the drop in classification performance on old classes, a phenomenon commonly called catastrophic forgetting. We propose in this paper a new method which exploits the availability of a large quantity of non-annotated images in addition to the annotated batches. These images are used to regularize the classifier and give the feature space a more stable structure. We demonstrate on several image data sets that our approach is able to improve the global performance of classifiers learned using an incremental learning protocol, even with annotated batches of small size.

Beyond the Deep Metric Learning: Enhance the Cross-Modal Matching with Adversarial Discriminative Domain Regularization

Li Ren, Kai Li, Liqiang Wang, Kien Hua

Auto-TLDR; Adversarial Discriminative Domain Regularization for Efficient Cross-Modal Matching

Abstract Slides Poster Similar

Matching information across image and text modalities is a fundamental challenge for many applications that involve both vision and natural language processing. The objective is to find efficient similarity metrics to compare the similarity between visual and textual information. Existing approaches mainly match the local visual objects and the sentence words in a shared space with attention mechanisms. The matching performance is still limited because the similarity computation is based on simple comparisons of the matching features, ignoring the characteristics of their distribution in the data. In this paper, we address this limitation with an efficient learning objective that considers the discriminative feature distributions between the visual objects and sentence words. Specifically, we propose a novel Adversarial Discriminative Domain Regularization (ADDR) learning framework, beyond the paradigm metric learning objective, to construct a set of discriminative data domains within each image-text pairs. Our approach can generally improve the learning efficiency and the performance of existing metrics learning frameworks by regulating the distribution of the hidden space between the matching pairs. The experimental results show that this new approach significantly improves the overall performance of several popular cross-modal matching techniques (SCAN, VSRN, BFAN) on the MS-COCO and Flickr30K benchmarks.

S2I-Bird: Sound-To-Image Generation of Bird Species Using Generative Adversarial Networks

Joo Yong Shim, Joongheon Kim, Jong-Kook Kim

Auto-TLDR; Generating bird images from sound using conditional generative adversarial networks

Abstract Slides Poster Similar

Generating images from sound is a challenging task. This paper proposes a novel deep learning model that generates bird images from their corresponding sound information. Our proposed model includes a sound encoder in order to extract suitable feature representations from audio recordings, and then it generates bird images that corresponds to its calls using conditional generative adversarial networks (GANs) with auxiliary classifiers. We demonstrate that our model produces better image generation results which outperforms other state-of-the-art methods in a similar context.

Progressive Cluster Purification for Unsupervised Feature Learning

Yifei Zhang, Chang Liu, Yu Zhou, Wei Wang, Weiping Wang, Qixiang Ye

Auto-TLDR; Progressive Cluster Purification for Unsupervised Feature Learning

Abstract Slides Poster Similar

In unsupervised feature learning, sample specificity based methods ignore the inter-class information, which deteriorates the discriminative capability of representation models. Clustering based methods are error-prone to explore the complete class boundary information due to the inevitable class inconsistent samples in each cluster. In this work, we propose a novel clustering based method, which, by iteratively excluding class inconsistent samples during progressive cluster formation, alleviates the impact of noise samples in a simple-yet-effective manner. Our approach, referred to as Progressive Cluster Purification (PCP), implements progressive clustering by gradually reducing the number of clusters during training, while the sizes of clusters continuously expand consistently with the growth of model representation capability. With a well-designed cluster purification mechanism, it further purifies clusters by filtering noise samples which facilitate the subsequent feature learning by utilizing the refined clusters as pseudo-labels. Experiments on commonly used benchmarks demonstrate that the proposed PCP improves baseline method with significant margins. Our code will be available at https://github.com/zhangyifei0115/PCP.

The Effect of Spectrogram Reconstruction on Automatic Music Transcription: An Alternative Approach to Improve Transcription Accuracy

Kin Wai Cheuk, Yin-Jyun Luo, Emmanouil Benetos, Herremans Dorien

Auto-TLDR; Exploring the effect of spectrogram reconstruction loss on automatic music transcription

Abstract Slides Similar

Most of the state-of-the-art automatic music transcription (AMT) models break down the main transcription task into sub-tasks such as onset prediction and offset prediction and train them with onset and offset labels. These predictions are then concatenated together and used as the input to train another model with the pitch labels to obtain the final transcription. We attempt to use only the pitch labels (together with spectrogram reconstruction loss) and explore how far this model can go without introducing supervised sub-tasks. In this paper, we do not aim at achieving state-of-the-art transcription accuracy, instead, we explore the effect that spectrogram reconstruction has on our AMT model. Our proposed model consists of two U-nets: the first U-net transcribes the spectrogram into a posteriorgram, and a second U-net transforms the posteriorgram back into a spectrogram. A reconstruction loss is applied between the original spectrogram and the reconstructed spectrogram to constrain the second U-net to focus only on reconstruction. We train our model on different datasets including MAPS, MAESTRO, and MusicNet. Our experiments show that adding the reconstruction loss can generally improve the note-level transcription accuracy when compared to the same model without the reconstruction part. Moreover, it can also boost the frame-level precision to be higher than the state-of-the-art models. The feature maps learned by our u-net contain gridlike structures (not present in the baseline model) which implies that with the present of reconstruction loss, the model is probably trying to count along both the time and frequency axis, resulting in a higher note-level transcription accuracy.

Attention-Based Deep Metric Learning for Near-Duplicate Video Retrieval

Kuan-Hsun Wang, Chia Chun Cheng, Yi-Ling Chen, Yale Song, Shang-Hong Lai

Auto-TLDR; Attention-based Deep Metric Learning for Near-duplicate Video Retrieval

Abstract Slides Similar

Near-duplicate video retrieval (NDVR) is an important and challenging problem due to the increasing amount of videos uploaded to the Internet. In this paper, we propose an attention-based deep metric learning method for NDVR. Our method is based on well-established principles: We leverage two-stream networks to combine RGB and optical flow features, and incorporate an attention module to effectively deal with distractor frames commonly observed in near duplicate videos. We further aggregate the features corresponding to multiple video segments to enhance the discriminative power. The whole system is trained using a deep metric learning objective with a Siamese architecture. Our experiments show that the attention module helps eliminate redundant and noisy frames, while focusing on visually relevant frames for solving NVDR. We evaluate our approach on recent large-scale NDVR datasets, CC_WEB_VIDEO, VCDB, FIVR and SVD. To demonstrate the generalization ability of our approach, we report results in both within- and cross-dataset settings, and show that the proposed method significantly outperforms state-of-the-art approaches.

Motion-Supervised Co-Part Segmentation

Aliaksandr Siarohin, Subhankar Roy, Stéphane Lathuiliere, Sergey Tulyakov, Elisa Ricci, Nicu Sebe

Auto-TLDR; Self-supervised Co-Part Segmentation Using Motion Information from Videos

Abstract Slides Similar

Recent co-part segmentation methods mostly operate in a supervised learning setting, which requires a large amount of annotated data for training. To overcome this limitation, we propose a self-supervised deep learning method for co-part segmentation. Differently from previous works, our approach develops the idea that motion information inferred from videos can be leveraged to discover meaningful object parts. To this end, our method relies on pairs of frames sampled from the same video. The network learns to predict part segments together with a representation of the motion between two frames, which permits reconstruction of the target image. Through extensive experimental evaluation on publicly available video sequences we demonstrate that our approach can produce improved segmentation maps with respect to previous self-supervised co-part segmentation approaches.