Showing papers by "Andrew Zisserman published in 2020"

End-to-End Learning of Visual Representations From Uncurated Instructional Videos

Abstract: Annotating videos is cumbersome, expensive and not scalable. Yet, many strong video models still rely on manually annotated data. With the recent introduction of the HowTo100M dataset, narrated videos now offer the possibility of learning video representations without manual supervision. In this work we propose a new learning approach, MIL-NCE, capable of addressing mis- alignments inherent in narrated videos. With this approach we are able to learn strong video representations from scratch, without the need for any manual annotation. We evaluate our representations on a wide range of four downstream tasks over eight datasets: action recognition (HMDB-51, UCF-101, Kinetics-700), text-to- video retrieval (YouCook2, MSR-VTT), action localization (YouTube-8M Segments, CrossTask) and action segmentation (COIN). Our method outperforms all published self-supervised approaches for these tasks as well as several fully supervised baselines.

Vggsound: A Large-Scale Audio-Visual Dataset

Abstract: Our goal is to collect a large-scale audio-visual dataset with low label noise from videos ‘in the wild’ using computer vision techniques. The resulting dataset can be used for training and evaluating audio recognition models. We make three contributions. First, we propose a scalable pipeline based on computer vision techniques to create an audio dataset from open-source media. Our pipeline involves obtaining videos from YouTube; using image classification algorithms to localize audio-visual correspondence; and filtering out ambient noise using audio verification. Second, we use this pipeline to curate the VGGSound dataset consisting of more than 200k videos for 300 audio classes. Third, we investigate various Convolutional Neural Network (CNN) architectures and aggregation approaches to establish audio recognition baselines for our new dataset. Compared to existing audio datasets, VGGSound ensures audio-visual correspondence and is collected under unconstrained conditions. Code and the dataset are available at http://www.robots.ox.ac.uk/~vgg/data/vggsound/.

Self-Supervised MultiModal Versatile Networks

Abstract: Videos are a rich source of multi-modal supervision. In this work, we learn representations using self-supervision by leveraging three modalities naturally present in videos: vision, audio and language. To this end, we introduce the notion of a multimodal versatile network -- a network that can ingest multiple modalities and whose representations enable downstream tasks in multiple modalities. In particular, we explore how best to combine the modalities, such that fine-grained representations of audio and vision can be maintained, whilst also integrating text into a common embedding. Driven by versatility, we also introduce a novel process of deflation, so that the networks can be effortlessly applied to the visual data in the form of video or a static image. We demonstrate how such networks trained on large collections of unlabelled video data can be applied on video, video-text, image and audio tasks. Equipped with these representations, we obtain state-of-the-art performance on multiple challenging benchmarks including UCF101, HMDB51 and ESC-50 when compared to previous self-supervised work.

Memory-augmented Dense Predictive Coding for Video Representation Learning

Abstract: The objective of this paper is self-supervised learning from video, in particular for representations for action recognition. We make the following contributions: (i) We propose a new architecture and learning framework Memory-augmented Dense Predictive Coding (MemDPC) for the task. It is trained with a predictive attention mechanism over the set of compressed memories, such that any future states can always be constructed by a convex combination of the condense representations, allowing to make multiple hypotheses efficiently. (ii) We investigate visual-only self-supervised video representation learning from RGB frames, or from unsupervised optical flow, or both. (iii) We thoroughly evaluate the quality of learnt representation on four different downstream tasks: action recognition, video retrieval, learning with scarce annotations, and unintentional action classification. In all cases, we demonstrate state-of-the-art or comparable performance over other approaches with orders of magnitude fewer training data.

Self-supervised Co-Training for Video Representation Learning

Abstract: The objective of this paper is visual-only self-supervised video representation learning. We make the following contributions: (i) we investigate the benefit of adding semantic-class positives to instance-based Info Noise Contrastive Estimation (InfoNCE) training, showing that this form of supervised contrastive learning leads to a clear improvement in performance; (ii) we propose a novel self-supervised co-training scheme to improve the popular infoNCE loss, exploiting the complementary information from different views, RGB streams and optical flow, of the same data source by using one view to obtain positive class samples for the other; (iii) we thoroughly evaluate the quality of the learnt representation on two different downstream tasks: action recognition and video retrieval. In both cases, the proposed approach demonstrates state-of-the-art or comparable performance with other self-supervised approaches, whilst being significantly more efficient to train, i.e. requiring far less training data to achieve similar performance.

CrossTransformers: spatially-aware few-shot transfer

Abstract: Given new tasks with very little data$-$such as new classes in a classification problem or a domain shift in the input$-$performance of modern vision systems degrades remarkably quickly. In this work, we illustrate how the neural network representations which underpin modern vision systems are subject to supervision collapse, whereby they lose any information that is not necessary for performing the training task, including information that may be necessary for transfer to new tasks or domains. We then propose two methods to mitigate this problem. First, we employ self-supervised learning to encourage general-purpose features that transfer better. Second, we propose a novel Transformer based neural network architecture called CrossTransformers, which can take a small number of labeled images and an unlabeled query, find coarse spatial correspondence between the query and the labeled images, and then infer class membership by computing distances between spatially-corresponding features. The result is a classifier that is more robust to task and domain shift, which we demonstrate via state-of-the-art performance on Meta-Dataset, a recent dataset for evaluating transfer from ImageNet to many other vision datasets.

Self-Supervised Learning of Audio-Visual Objects from Video

Abstract: Our objective is to transform a video into a set of discrete audio-visual objects using self-supervised learning. To this end, we introduce a model that uses attention to localize and group sound sources, and optical flow to aggregate information over time. We demonstrate the effectiveness of the audio-visual object embeddings that our model learns by using them for four downstream speech-oriented tasks: (a) multi-speaker sound source separation, (b) localizing and tracking speakers, (c) correcting misaligned audio-visual data, and (d) active speaker detection. Using our representation, these tasks can be solved entirely by training on unlabeled video, without the aid of object detectors. We also demonstrate the generality of our method by applying it to non-human speakers, including cartoons and puppets.Our model significantly outperforms other self-supervised approaches, and obtains performance competitive with methods that use supervised face detection.

Spot the Conversation: Speaker Diarisation in the Wild.

Abstract: The goal of this paper is speaker diarisation of videos collected 'in the wild'. We make three key contributions. First, we propose an automatic audio-visual diarisation method for YouTube videos. Our method consists of active speaker detection using audio-visual methods and speaker verification using self-enrolled speaker models. Second, we integrate our method into a semi-automatic dataset creation pipeline which significantly reduces the number of hours required to annotate videos with diarisation labels. Finally, we use this pipeline to create a large-scale diarisation dataset called VoxConverse, collected from 'in the wild' videos, which we will release publicly to the research community. Our dataset consists of overlapping speech, a large and diverse speaker pool, and challenging background conditions.

Self-supervised Co-training for Video Representation Learning

Abstract: The objective of this paper is visual-only self-supervised video representation learning. We make the following contributions: (i) we investigate the benefit of adding semantic-class positives to instance-based Info Noise Contrastive Estimation (InfoNCE) training, showing that this form of supervised contrastive learning leads to a clear improvement in performance; (ii) we propose a novel self-supervised co-training scheme to improve the popular infoNCE loss, exploiting the complementary information from different views, RGB streams and optical flow, of the same data source by using one view to obtain positive class samples for the other; (iii) we thoroughly evaluate the quality of the learnt representation on two different downstream tasks: action recognition and video retrieval. In both cases, the proposed approach demonstrates state-of-the-art or comparable performance with other self-supervised approaches, whilst being significantly more efficient to train, i.e. requiring far less training data to achieve similar performance.

Smooth-AP: Smoothing the Path Towards Large-Scale Image Retrieval.

Abstract: Optimising a ranking-based metric, such as Average Precision (AP), is notoriously challenging due to the fact that it is non-differentiable, and hence cannot be optimised directly using gradient-descent methods. To this end, we introduce an objective that optimises instead a smoothed approximation of AP, coined Smooth-AP. Smooth-AP is a plug-and-play objective function that allows for end-to-end training of deep networks with a simple and elegant implementation. We also present an analysis for why directly optimising the ranking based metric of AP offers benefits over other deep metric learning losses.

BSL-1K: Scaling Up Co-articulated Sign Language Recognition Using Mouthing Cues

Abstract: Recent progress in fine-grained gesture and action classification, and machine translation, point to the possibility of automated sign language recognition becoming a reality. A key stumbling block in making progress towards this goal is a lack of appropriate training data, stemming from the high complexity of sign annotation and a limited supply of qualified annotators. In this work, we introduce a new scalable approach to data collection for sign recognition in continuous videos. We make use of weakly-aligned subtitles for broadcast footage together with a keyword spotting method to automatically localise sign-instances for a vocabulary of 1,000 signs in 1,000 h of video. We make the following contributions: (1) We show how to use mouthing cues from signers to obtain high-quality annotations from video data—the result is the BSL-1K dataset, a collection of British Sign Language (BSL) signs of unprecedented scale; (2) We show that we can use BSL-1K to train strong sign recognition models for co-articulated signs in BSL and that these models additionally form excellent pretraining for other sign languages and benchmarks—we exceed the state of the art on both the MSASL and WLASL benchmarks. Finally, (3) we propose new large-scale evaluation sets for the tasks of sign recognition and sign spotting and provide baselines which we hope will serve to stimulate research in this area.

Self-Supervised Learning of Audio-Visual Objects from Video

Abstract: Our objective is to transform a video into a set of discrete audio-visual objects using self-supervised learning. To this end, we introduce a model that uses attention to localize and group sound sources, and optical flow to aggregate information over time. We demonstrate the effectiveness of the audio-visual object embeddings that our model learns by using them for four downstream speech-oriented tasks: (a) multi-speaker sound source separation, (b) localizing and tracking speakers, (c) correcting misaligned audio-visual data, and (d) active speaker detection. Using our representation, these tasks can be solved entirely by training on unlabeled video, without the aid of object detectors. We also demonstrate the generality of our method by applying it to non-human speakers, including cartoons and puppets. Our model significantly outperforms other self-supervised approaches, and obtains performance competitive with methods that use supervised face detection.

Automatically Discovering and Learning New Visual Categories with Ranking Statistics

Abstract: We tackle the problem of discovering novel classes in an image collection given labelled examples of other classes. This setting is similar to semi-supervised learning, but significantly harder because there are no labelled examples for the new classes. The challenge, then, is to leverage the information contained in the labelled images in order to learn a general-purpose clustering model and use the latter to identify the new classes in the unlabelled data. In this work we address this problem by combining three ideas: (1) we suggest that the common approach of bootstrapping an image representation using the labeled data only introduces an unwanted bias, and that this can be avoided by using self-supervised learning to train the representation from scratch on the union of labelled and unlabelled data; (2) we use rank statistics to transfer the model's knowledge of the labelled classes to the problem of clustering the unlabelled images; and, (3) we train the data representation by optimizing a joint objective function on the labelled and unlabelled subsets of the data, improving both the supervised classification of the labelled data, and the clustering of the unlabelled data. We evaluate our approach on standard classification benchmarks and outperform current methods for novel category discovery by a significant margin.

The AVA-Kinetics Localized Human Actions Video Dataset.

Abstract: This paper describes the AVA-Kinetics localized human actions video dataset. The dataset is collected by annotating videos from the Kinetics-700 dataset using the AVA annotation protocol, and extending the original AVA dataset with these new AVA annotated Kinetics clips. The dataset contains over 230k clips annotated with the 80 AVA action classes for each of the humans in key-frames. We describe the annotation process and provide statistics about the new dataset. We also include a baseline evaluation using the Video Action Transformer Network on the AVA-Kinetics dataset, demonstrating improved performance for action classification on the AVA test set. The dataset can be downloaded from this https URL

Amplifying Key Cues for Human-Object-Interaction Detection

Abstract: Human-object interaction (HOI) detection aims to detect and recognise how people interact with the objects that surround them. This is challenging as different interaction categories are often distinguished only by very subtle visual differences in the scene. In this paper we introduce two methods to amplify key cues in the image, and also a method to combine these and other cues when considering the interaction between a human and an object. First, we introduce an encoding mechanism for representing the fine-grained spatial layout of the human and object (a subtle cue) and also semantic context (a cue, represented by text embeddings of surrounding objects). Second, we use plausible future movements of humans and objects as a cue to constrain the space of possible interactions. Third, we use a gate and memory architecture as a fusion module to combine the cues. We demonstrate that these three improvements lead to a performance which exceeds prior HOI methods across standard benchmarks by a considerable margin.

Disentangled Speech Embeddings Using Cross-Modal Self-Supervision

Abstract: The objective of this paper is to learn representations of speaker identity without access to manually annotated data. To do so, we develop a self-supervised learning objective that exploits the natural cross-modal synchrony between faces and audio in video. The key idea behind our approach is to tease apart—without annotation—the representations of linguistic content and speaker identity. We construct a two-stream architecture which: (1) shares low-level features common to both representations; and (2) provides a natural mechanism for explicitly disentangling these factors, offering the potential for greater generalisation to novel combinations of content and identity and ultimately producing speaker identity representations that are more robust.We train our method on a large-scale audio-visual dataset of talking heads ‘in the wild’, and demonstrate its efficacy by evaluating the learned speaker representations for standard speaker recognition performance.

ASR is All You Need: Cross-Modal Distillation for Lip Reading

Abstract: The goal of this work is to train strong models for visual speech recognition without requiring human annotated ground truth data. We achieve this by distilling from an Automatic Speech Recognition (ASR) model that has been trained on a large-scale audio-only corpus. We use a cross-modal distillation method that combines Connectionist Temporal Classification (CTC) with a frame-wise cross-entropy loss. Our contributions are fourfold: (i) we show that ground truth transcriptions are not necessary to train a lip reading system; (ii) we show how arbitrary amounts of unlabelled video data can be leveraged to improve performance; (iii) we demonstrate that distillation significantly speeds up training; and, (iv) we obtain state-of-the-art results on the challenging LRS2 and LRS3 datasets for training only on publicly available data.

Condensed Movies: Story Based Retrieval with Contextual Embeddings

Abstract: Our objective in this work is long range understanding of the narrative structure of movies. Instead of considering the entire movie, we propose to learn from the ‘key scenes’ of the movie, providing a condensed look at the full storyline. To this end, we make the following three contributions: (i) We create the Condensed Movies Dataset (CMD) consisting of the key scenes from over 3 K movies: each key scene is accompanied by a high level semantic description of the scene, character face-tracks, and metadata about the movie. The dataset is scalable, obtained automatically from YouTube, and is freely available for anybody to download and use. It is also an order of magnitude larger than existing movie datasets in the number of movies; (ii) We provide a deep network baseline for text-to-video retrieval on our dataset, combining character, speech and visual cues into a single video embedding; and finally (iii) We demonstrate how the addition of context from other video clips improves retrieval performance.

Memory-Augmented Dense Predictive Coding for Video Representation Learning

Abstract: The objective of this paper is self-supervised learning from video, in particular for representations for action recognition. We make the following contributions: (i) We propose a new architecture and learning framework Memory-augmented Dense Predictive Coding (MemDPC) for the task. It is trained with a predictive attention mechanism over the set of compressed memories, such that any future states can always be constructed by a convex combination of the condensed representations, allowing to make multiple hypotheses efficiently. (ii) We investigate visual-only self-supervised video representation learning from RGB frames, or from unsupervised optical flow, or both. (iii) We thoroughly evaluate the quality of the learnt representation on four different downstream tasks: action recognition, video retrieval, learning with scarce annotations, and unintentional action classification. In all cases, we demonstrate state-of-the-art or comparable performance over other approaches with orders of magnitude fewer training data.

VoxSRC 2020: The Second VoxCeleb Speaker Recognition Challenge.

Abstract: We held the second installment of the VoxCeleb Speaker Recognition Challenge in conjunction with Interspeech 2020. The goal of this challenge was to assess how well current speaker recognition technology is able to diarise and recognize speakers in unconstrained or `in the wild' data. It consisted of: (i) a publicly available speaker recognition and diarisation dataset from YouTube videos together with ground truth annotation and standardised evaluation software; and (ii) a virtual public challenge and workshop held at Interspeech 2020. This paper outlines the challenge, and describes the baselines, methods used, and results. We conclude with a discussion of the progress over the first installment of the challenge.

CrossTransformers: spatially-aware few-shot transfer

Abstract: Given new tasks with very little data$-$such as new classes in a classification problem or a domain shift in the input$-$performance of modern vision systems degrades remarkably quickly. In this work, we illustrate how the neural network representations which underpin modern vision systems are subject to supervision collapse, whereby they lose any information that is not necessary for performing the training task, including information that may be necessary for transfer to new tasks or domains. We then propose two methods to mitigate this problem. First, we employ self-supervised learning to encourage general-purpose features that transfer better. Second, we propose a novel Transformer based neural network architecture called CrossTransformers, which can take a small number of labeled images and an unlabeled query, find coarse spatial correspondence between the query and the labeled images, and then infer class membership by computing distances between spatially-corresponding features. The result is a classifier that is more robust to task and domain shift, which we demonstrate via state-of-the-art performance on Meta-Dataset, a recent dataset for evaluating transfer from ImageNet to many other vision datasets.

Counting Out Time: Class Agnostic Video Repetition Counting in the Wild

Abstract: We present an approach for estimating the period with which an action is repeated in a video. The crux of the approach lies in constraining the period prediction module to use temporal self-similarity as an intermediate representation bottleneck that allows generalization to unseen repetitions in videos in the wild. We train this model, called RepNet, with a synthetic dataset that is generated from a large unlabeled video collection by sampling short clips of varying lengths and repeating them with different periods and counts. This combination of synthetic data and a powerful yet constrained model, allows us to predict periods in a class-agnostic fashion. Our model substantially exceeds the state of the art performance on existing periodicity (PERTUBE) and repetition counting (QUVA) benchmarks. We also collect a new challenging dataset called Countix (~90 times larger than existing datasets) which captures the challenges of repetition counting in real-world videos. Project webpage: https://sites.google.com/view/repnet .

Training Neural Networks for and by Interpolation

Abstract: The majority of modern deep learning models are able to interpolate the data: the empirical loss can be driven near zero on all samples simultaneously. In this work, we explicitly exploit this interpolation property for the design of a new optimization algorithm for deep learning. Specifically, we use it to compute an adaptive learning-rate given a stochastic gradient direction. This results in the Adaptive Learning-rates for Interpolation with Gradients (ALI-G) algorithm. ALI-G retains the advantages of SGD, which are low computational cost and provable convergence in the convex setting. But unlike SGD, the learning-rate of ALI-G can be computed inexpensively in closed-form and does not require a manual schedule. We provide a detailed analysis of ALI-G in the stochastic convex setting with explicit convergence rates. In order to obtain good empirical performance in deep learning, we extend the algorithm to use a maximal learning-rate, which gives a single hyper-parameter to tune. We show that employing such a maximal learning-rate has an intuitive proximal interpretation and preserves all convergence guarantees. We provide experiments on a variety of architectures and tasks: (i) learning a differentiable neural computer; (ii) training a wide residual network on the SVHN data set; (iii) training a Bi-LSTM on the SNLI data set; and (iv) training wide residual networks and densely connected networks on the CIFAR data sets. We empirically show that ALI-G outperforms adaptive gradient methods such as Adam, and provides comparable performance with SGD, although SGD benefits from manual learning rate schedules. We release PyTorch and Tensorflow implementations of ALI-G as standalone optimizers that can be used as a drop-in replacement in existing code (code available at this https URL ).

Self-Supervised MultiModal Versatile Networks

Abstract: Videos are a rich source of multi-modal supervision In this work, we learn representations using self-supervision by leveraging three modalities naturally present in videos: vision, audio and language To this end, we introduce the notion of a multimodal versatile network -- a network that can ingest multiple modalities and whose representations enable downstream tasks in multiple modalities In particular, we explore how best to combine the modalities, such that fine-grained representations of audio and vision can be maintained, whilst also integrating text into a common embedding Driven by versatility, we also introduce a novel process of deflation, so that the networks can be effortlessly applied to the visual data in the form of video or a static image We demonstrate how such networks trained on large collections of unlabelled video data can be applied on video, video-text, image and audio tasks Equipped with these representations, we obtain state-of-the-art performance on multiple challenging benchmarks including UCF101, HMDB51 and ESC-50 when compared to previous self-supervised work

Betrayed by Motion: Camouflaged Object Discovery via Motion Segmentation

Abstract: The objective of this paper is to design a computational architecture that discovers camouflaged objects in videos, specifically by exploiting motion information to perform object segmentation. We make the following three contributions: (i) We propose a novel architecture that consists of two essential components for breaking camouflage, namely, a differentiable registration module to align consecutive frames based on the background, which effectively emphasises the object boundary in the difference image, and a motion segmentation module with memory that discovers the moving objects, while maintaining the object permanence even when motion is absent at some point. (ii) We collect the first large-scale Moving Camouflaged Animals (MoCA) video dataset, which consists of over 140 clips across a diverse range of animals (67 categories). (iii) We demonstrate the effectiveness of the proposed model on MoCA, and achieve competitive performance on the unsupervised segmentation protocol on DAVIS2016 by only relying on motion.

Speech2Action: Cross-modal Supervision for Action Recognition

Abstract: Is it possible to guess human action from dialogue alone? In this work we investigate the link between spoken words and actions in movies. We note that movie screenplays describe actions, as well as contain the speech of characters and hence can be used to learn this correlation with no additional supervision. We train a BERT-based Speech2Action classifier on over a thousand movie screenplays, to predict action labels from transcribed speech segments. We then apply this model to the speech segments of a large unlabelled movie corpus (188M speech segments from 288K movies). Using the predictions of this model, we obtain weak action labels for over 800K video clips. By training on these video clips, we demonstrate superior action recognition performance on standard action recognition benchmarks, without using a single manually labelled action example.

Semi-Supervised Learning with Scarce Annotations

Abstract: While semi-supervised learning (SSL) algorithms provide an efficient way to make use of both labelled and unlabelled data, they generally struggle when the number of annotated samples is very small In this work, we consider the problem of SSL multi-class classification with very few labelled instances We introduce two key ideas The first is a simple but effective one: we leverage the power of transfer learning among different tasks and self-supervision to initialize a good representation of the data without making use of any label The second idea is a new algorithm for SSL that can exploit well such a pre-trained representation The algorithm works by alternating two phases, one fitting the labelled points and one fitting the unlabelled ones, with carefully-controlled information flow between them The benefits are greatly reducing overfitting of the labelled data and avoiding issue with balancing labelled and unlabelled losses during training We show empirically that this method can successfully train competitive models with as few as 10 labelled data points per class More in general, we show that the idea of bootstrapping features using self-supervised learning always improves SSL on standard benchmarks We show that our algorithm works increasingly well compared to other methods when refining from other tasks or datasets

Counting Out Time: Class Agnostic Video Repetition Counting in the Wild

Abstract: We present an approach for estimating the period with which an action is repeated in a video The crux of the approach lies in constraining the period prediction module to use temporal self-similarity as an intermediate representation bottleneck that allows generalization to unseen repetitions in videos in the wild We train this model, called Repnet, with a synthetic dataset that is generated from a large unlabeled video collection by sampling short clips of varying lengths and repeating them with different periods and counts This combination of synthetic data and a powerful yet constrained model, allows us to predict periods in a class-agnostic fashion Our model substantially exceeds the state of the art performance on existing periodicity (PERTUBE) and repetition counting (QUVA) benchmarks We also collect a new challenging dataset called Countix (~90 times larger than existing datasets) which captures the challenges of repetition counting in real-world videos Project webpage: this https URL

Speech2Action: Cross-Modal Supervision for Action Recognition

Abstract: Is it possible to guess human action from dialogue alone? In this work we investigate the link between spoken words and actions in movies. We note that movie screenplays describe actions, as well as contain the speech of characters and hence can be used to learn this correlation with no additional supervision. We train a BERT-based Speech2Action classifier on over a thousand movie screenplays, to predict action labels from transcribed speech segments. We then apply this model to the speech segments of a large unlabelled movie corpus (188M speech segments from 288K movies). Using the predictions of this model, we obtain weak action labels for over 800K video clips. By training on these video clips, we demonstrate superior action recognition performance on standard action recognition benchmarks, without using a single manually labelled action example.