We introduce ED-VTG, a method for fine-grained video temporal grounding utilizing multi-modal large language models. Our approach harnesses the capabilities of multimodal LLMs to jointly process text and video, in order to effectively localize natural language queries in videos through a two-stage process. Rather than being directly grounded, language queries are initially transformed into enriched sentences that incorporate missing details and cues to aid in grounding. In the second stage, these enriched queries are grounded, using a lightweight decoder, which specializes at predicting accurate boundaries conditioned on contextualized representations of the enriched queries. To mitigate noise and reduce the impact of hallucinations, our model is trained with a multiple-instance-learning objective that dynamically selects the optimal version of the query for each training sample. We demonstrate state-of-the-art results across various benchmarks in temporal video grounding and paragraph grounding settings. Experiments reveal that our method significantly outperforms all previously proposed LLM-based temporal grounding approaches and is either superior or comparable to specialized models, while maintaining a clear advantage against them in zero-shot evaluation scenarios.

Vision-language models are integral to computer vision research, yet many high-performing models remain closed-source, obscuring their data, design and training recipe. The research community has responded by using distillation from black-box models to label training data, achieving strong benchmark results, at the cost of measurable scientific progress. However, without knowing the details of the teacher model and its data sources, scientific progress remains difficult to measure. In this paper, we study building a Perception Language Model (PLM) in a fully open and reproducible framework for transparent research in image and video understanding. We analyze standard training pipelines without distillation from proprietary models and explore large-scale synthetic data to identify critical data gaps, particularly in detailed video understanding. To bridge these gaps, we release 2.8M human-labeled instances of fine-grained video question-answer pairs and spatio-temporally grounded video captions. Additionally, we introduce PLM–VideoBench, a suite for evaluating challenging video understanding tasks focusing on the ability to reason about ''what'', ''where'', ''when'', and ''how'' of a video. We make our work fully reproducible by providing data, training recipes, code & models.

We present Ego-Exo4D, a diverse, large-scale multimodal multiview video dataset and benchmark challenge. Ego-Exo4D centers around simultaneously-captured egocentric and exocentric video of skilled human activities (e.g., sports, music, dance, bike repair). 740 participants from 13 cities worldwide performed these activities in 123 different natural scene contexts, yielding long-form captures from 1 to 42 minutes each and 1,286 hours of video combined. The multimodal nature of the dataset is unprecedented: the video is accompanied by multichannel audio, eye gaze, 3D point clouds, camera poses, IMU, and multiple paired language descriptions—including a novel “expert commentary” done by coaches and teachers and tailored to the skilled-activity domain. To push the frontier of first-person video understanding of skilled human activity, we also present a suite of benchmark tasks and their annotations, including fine-grained activity understanding, proficiency estimation, cross-view translation, and 3D hand/body pose. All resources are open sourced to fuel new research in the community. https://ego-exo4d-data.org/

We present Ego-Exo4D, a diverse, large-scale multi-modal multiview video dataset and benchmark challenge. Ego-Exo4D centers around simultaneously-captured ego-centric and exocentric video of skilled human activities (e.g., sports, music, dance, bike repair). 740 participants from 13 cities worldwide performed these activities in 123 different natural scene contexts, yielding long-form captures from 1 to 42 minutes each and 1,286 hours of video combined. The multimodal nature of the dataset is un-precedented: the video is accompanied by multichannel audio, eye gaze, 3D point clouds, camera poses, IMU, and multiple paired language descriptions-including a novel “expert commentary” done by coaches and teachers and tailored to the skilled-activity domain. To push the frontier of first-person video understanding of skilled human activity, we also present a suite of benchmark tasks and their annotations, including fine-grained activity understanding, proficiency estimation, cross-view translation, and 3D hand/body pose. All resources are open sourced to fuel new research in the community.

We introduce MusicFlow, a cascaded text-to-music generation model based on flow matching. Based on self-supervised representations to bridge between text descriptions and music audios, we construct two flow matching networks to model the conditional distribution of semantic and acoustic features. Additionally, we leverage masked prediction as the training objective, enabling the model to generalize to other tasks such as music infilling and continuation in a zero-shot manner. Experiments on MusicCaps reveal that the music generated by MusicFlow exhibits superior quality and text coherence despite being over $2\sim5$ times smaller and requiring $5$ times fewer iterative steps. Simultaneously, the model can perform other music generation tasks and achieves competitive performance in music infilling and continuation.

In this paper we present an approach for localizing steps of procedural activities in narrated how-to videos. To deal with the scarcity of labeled data at scale, we source the step descriptions from a language knowledge base (wikiHow) containing instructional articles for a large variety of procedural tasks. Without any form of manual supervision, our model learns to temporally ground the steps of procedural articles in how-to videos by matching three modalities: frames, narrations, and step descriptions. Specifically, our method aligns steps to video by fusing information from two distinct pathways: i) direct alignment of step descriptions to frames, ii) indirect alignment obtained by composing steps-to-narrations with narrations-to-video correspondences. Notably, our approach performs global temporal grounding of all steps in an article at once by exploiting order information, and is trained with step pseudo-labels which are iteratively refined and aggressively filtered. In order to validate our model we introduce a new benchmark – HT-Step – obtained by manually annotating a 124-hour subset of HowTo100M1with steps sourced from wikiHow articles. Experiments on this benchmark as well as zero-shot evaluations on CrossTask demonstrate that our multi-modality alignment yields dramatic gains over several baselines and prior works. Finally, we show that our inner module for matching narration-to-video outperforms by a large margin the state of the art on the HTM-Align narration-video alignment benchmark.

We introduce HT-Step, a large-scale dataset containing temporal annotations of instructional article steps in cooking videos. It includes 122k segment-level annotations over 20k narrated videos (approximately 2.3k hours) of the HowTo100M dataset.Each annotation provides a temporal interval, and a categorical step label from a taxonomy of 4,958 unique steps automatically mined from wikiHow articles which include rich descriptions of each step.Our dataset significantly surpasses existing labeled step datasets in terms of scale, number of tasks, and richness of natural language step descriptions. Based on these annotations, we introduce a strongly supervised benchmark for aligning instructional articles with how-to videos and present a comprehensive evaluation of baseline methods for this task.By publicly releasing these annotations and defining rigorous evaluation protocols and metrics,we hope to significantly accelerate research in the field of procedural activity understanding.

Procedural activity understanding requires perceiving human actions in terms of a broader task, where multiple keysteps are performed in sequence across a long video to reach a final goal state---such as the steps of a recipe or the steps of a DIY fix-it task. Prior work largely treats keystep recognition in isolation of this broader structure, or else rigidly confines keysteps to align with a particular sequential script. We propose discovering a task graph automatically from how-to videos to represent probabilistically how people tend to execute keysteps, then leverage this graph to regularize keystep recognition in novel videos. On multiple datasets of real-world instructional video, we show the impact: more reliable zero-shot keystep localization and improved video representation learning, exceeding the state of the art.

Humans are remarkably sensitive to the alignment of visual events with other stimuli, which makes synchronization one of the hardest tasks in video editing. A key observation of our work is that most of the alignment we do involves salient localizable events that occur sparsely in time. By learning how to recognize these events, we can greatly reduce the space of possible synchronizations that an editor or algorithm has to consider. Furthermore, by learning descriptors of these events that capture additional properties of visible motion, we can build active tools that adapt their notion of eventfulness to a given task as they are being used. Rather than learning an automatic solution to one specific problem, our goal is to make a much broader class of interactive alignment tasks significantly easier and less time-consuming. We show that a suitable visual event descriptor can be learned entirely from stochastically-generated synthetic video. We then demonstrate the usefulness of learned and adaptive eventfulness by integrating it in novel interactive tools for applications including audio-driven time warping of video and the extraction and application of sound effects across different videos.

We tackle the problem of learning object detectors without supervision. Differently from weakly-supervised object detection, we do not assume image-level class labels. Instead, we extract a supervisory signal from audio-visual data, using the audio component to “teach” the object detector. While this problem is related to sound source localisation, it is considerably harder because the detector must classify the objects by type, enumerate each instance of the object, and do so even when the object is silent. We tackle this problem by first designing a self-supervised framework with a contrastive objective that jointly learns to classify and localise objects. Then, without using any supervision, we simply use these self-supervised labels and boxes to train an image-based object detector. With this, we outperform previous unsupervised and weakly-supervised detectors for the task of object detection and sound source localization. We also show that we can align this detector to ground-truth classes with as little as one label per pseudo-class, and show how our method can learn to detect generic objects that go beyond instruments, such as airplanes and cats.

The goal of this paper is to learn strong lip reading models that can recognise speech in silent videos. Most prior works deal with the open-set visual speech recognition problem by adapting existing automatic speech recognition techniques on top of trivially pooled visual features. Instead, in this paper, we focus on the unique challenges encountered in lip reading and propose tailored solutions. To this end, we make the following contributions: (1) we propose an attention-based pooling mechanism to aggregate visual speech representations; (2) we use sub-word units for lip reading for the first time and show that this allows us to better model the ambiguities of the task; (3) we propose a model for Visual Speech Detection (VSD), trained on top of the lip reading network. Following the above, we obtain state-of-the-art results on the challenging LRS2 and LRS3 benchmarks when training on public datasets, and even surpass models trained on large-scale industrial datasets by using an order of magnitude less data. Our best model achieves 22.6% word error rate on the LRS2 dataset, a performance unprecedented for lip reading models, significantly reducing the performance gap between lip reading and automatic speech recognition. Moreover, on the AVA-ActiveSpeaker benchmark, our VSD model surpasses all visual-only baselines and even outperforms several recent audio-visual methods.

The goal of this paper is speech separation and enhancement in multi-speaker and noisy environments using a combination of different modalities. Previous works have shown good performance when conditioning on temporal or static visual evidence such as synchronised lip movements or face identity. In this paper, we present a unified framework for multi-modal speech separation and enhancement based on synchronous or asynchronous cues. To that end we make the following contributions: (i) we design a modern Transformer-based architecture tailored to fuse different modalities to solve the speech separation task in the raw waveform domain; (ii) we propose conditioning on the textual content of a sentence alone or in combination with visual information; (iii) we demonstrate the robustness of our model to audio-visual synchronisation offsets; and, (iv) we obtain state-of-the-art performance on the well-established benchmark datasets LRS2 and LRS3.

Abstract The focus of this work issign spotting–given a video of an isolated sign, our task is to identifywhetherandwhereit has been signed in a continuous, co-articulated sign language video. To achieve this sign spotting task, we train a model using multiple types of available supervision by: (1)watchingexisting footage which is sparsely labelled using mouthing cues; (2)readingassociated subtitles (readily available translations of the signed content) which provide additionalweak-supervision; (3)looking upwords (for which no co-articulated labelled examples are available) in visual sign language dictionaries to enable novel sign spotting. These three tasks are integrated into a unified learning framework using the principles of Noise Contrastive Estimation and Multiple Instance Learning. We validate the effectiveness of our approach on low-shot sign spotting benchmarks. In addition, we contribute a machine-readable British Sign Language (BSL) dictionary dataset of isolated signs,BslDict, to facilitate study of this task. The dataset, models and code are available at our project page.

The goal of this work is to recognise phrases and sentences being spoken by a talking face, with or without the audio. Unlike previous works that have focussed on recognising a limited number of words or phrases, we tackle lip reading as an open-world problem - unconstrained natural language sentences, and in the wild videos. Our key contributions are: (1) we compare two models for lip reading, one using a CTC loss, and the other using a sequence-to-sequence loss. Both models are built on top of the transformer self-attention architecture; (2) we investigate to what extent lip reading is complementary to audio speech recognition, especially when the audio signal is noisy; (3) we introduce and publicly release a new dataset for audio-visual speech recognition, LRS2-BBC, consisting of thousands of natural sentences from British television. The models that we train surpass the performance of all previous work on a lip reading benchmark dataset by a significant margin.

The objective of this work is to localize sound sources that are visible in a video without using manual annotations. Our key technical contribution is to show that, by training the network to explicitly discriminate challenging image fragments, even for images that do contain the object emitting the sound, we can significantly boost the localization performance. We do so elegantly by introducing a mechanism to mine hard samples and add them to a contrastive learning formulation automatically. We show that our algorithm achieves state-of-the-art performance on the popular Flickr SoundNet dataset. Furthermore, we introduce the VGG-Sound Source (VGG-SS) benchmark, a new set of annotations for the recently-introduced VGG-Sound dataset, where the sound sources visible in each video clip are explicitly marked with bounding box annotations. This dataset is 20 times larger than analogous existing ones, contains 5K videos spanning over 200 categories, and, differently from Flickr SoundNet, is video-based. On VGG-SS, we also show that our algorithm achieves state-of-the-art performance against several baselines. Code and datasets can be found at http://www.robots.ox.ac.uk/˜vgg/research/lvs/.

The objective of this work is to annotate sign instances across a broad vocabulary in continuous sign language. We train a Transformer model to ingest a continuous signing stream and output a sequence of written tokens on a large-scale collection of signing footage with weakly-aligned subtitles. We show that through this training it acquires the ability to attend to a large vocabulary of sign instances in the input sequence, enabling their localisation. Our contributions are as follows: (1) we demonstrate the ability to leverage large quantities of continuous signing videos with weakly-aligned subtitles to localise signs in continuous sign language; (2) we employ the learned attention to automatically generate hundreds of thousands of annotations for a large sign vocabulary; (3) we collect a set of 37K manually verified sign instances across a vocabulary of 950 sign classes to support our study of sign language recognition; (4) by training on the newly annotated data from our method, we outperform the prior state of the art on the BSL-1K sign language recognition benchmark.

In this work, we propose a framework to collect a large-scale, diverse sign language dataset that can be used to train automatic sign language recognition models.The first contribution of this work is SDTrack, a generic method for signer tracking and diarisation in the wild. Our second contribution is SeeHear, a dataset of 90 hours of British Sign Language (BSL) content featuring more than 1000 signers, and including interviews, monologues and debates. Using SDTrack, the SeeHear dataset is annotated with 35K active signing tracks, with corresponding signer identities and subtitles, and 40K automatically localised sign labels. As a third contribution, we provide benchmarks for signer diarisation and sign recognition on SeeHear.

The goal of this work is to temporally align asynchronous subtitles in sign language videos. In particular, we focus on sign-language interpreted TV broadcast data comprising (i) a video of continuous signing, and (ii) subtitles corresponding to the audio content. Previous work exploiting such weakly-aligned data only considered finding keyword-sign correspondences, whereas we aim to localise a complete subtitle text in continuous signing. We propose a Transformer architecture tailored for this task, which we train on manually annotated alignments covering over 15K subtitles that span 17.7 hours of video. We use BERT subtitle embeddings and CNN video representations learned for sign recognition to encode the two signals, which interact through a series of attention layers. Our model outputs frame-level predictions, i.e., for each video frame, whether it belongs to the queried subtitle or not. Through extensive evaluations, we show substantial improvements over existing alignment baselines that do not make use of subtitle text embeddings for learning. Our automatic alignment model opens up possibilities for advancing machine translation of sign languages via providing continuously synchronized video-text data.

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.

The goal of this work is to train models that can identify a spoken language just by interpreting the speaker's lip movements.Our contributions are the following: (i) we show that models can learn to discriminate among 14 different languages using only visual speech information; (ii) we compare different designs in sequence modelling and utterance-level aggregation in order to determine the best architecture for this task; (iii) we investigate the factors that contribute discriminative cues and show that our model indeed solves the problem by finding temporal patterns in mouth movements and not by exploiting spurious correlations.We demonstrate this further by evaluating our models on challenging examples from bilingual speakers.

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.

Our objective is an audio-visual model for separating a single speaker from a mixture of sounds such as other speakers and background noise. Moreover, we wish to hear the speaker even when the visual cues are temporarily absent due to occlusion. To this end we introduce a deep audio-visual speech enhancement network that is able to separate a speaker's voice by conditioning on both the speaker's lip movements and/or a representation of their voice. The voice representation can be obtained by either (i) enrollment, or (ii) by self-enrollment -- learning the representation on-the-fly given sufficient unobstructed visual input. The model is trained by blending audios, and by introducing artificial occlusions around the mouth region that prevent the visual modality from dominating. The method is speaker-independent, and we demonstrate it on real examples of speakers unheard (and unseen) during training. The method also improves over previous models in particular for cases of occlusion in the visual modality.

Many real-world problems, such as network packet routing and the coordination of autonomous vehicles, are naturally modelled as cooperative multi-agent systems. There is a great need for new reinforcement learning methods that can efficiently learn decentralised policies for such systems. To this end, we propose a new multi-agent actor-critic method called counterfactual multi-agent (COMA) policy gradients. COMA uses a centralised critic to estimate the Q-function and decentralised actors to optimise the agents' policies. In addition, to address the challenges of multi-agent credit assignment, it uses a counterfactual baseline that marginalises out a single agent's action, while keeping the other agents' actions fixed. COMA also uses a critic representation that allows the counterfactual baseline to be computed efficiently in a single forward pass. We evaluate COMA in the testbed of StarCraft unit micromanagement, using a decentralised variant with significant partial observability. COMA significantly improves average performance over other multi-agent actor-critic methods in this setting, and the best performing agents are competitive with state-of-the-art centralised controllers that get access to the full state.

Our goal is to isolate individual speakers from multi-talker simultaneous speech in videos. Existing works in this area have focussed on trying to separate utterances from known speakers in controlled environments. In this paper, we propose a deep audio-visual speech enhancement network that is able to separate a speaker's voice given lip regions in the corresponding video, by predicting both the magnitude and the phase of the target signal. The method is applicable to speakers unheard and unseen during training, and for unconstrained environments. We demonstrate strong quantitative and qualitative results, isolating extremely challenging real-world examples.

The goal of this paper is to develop state-of-the-art models for lip reading -- visual speech recognition. We develop three architectures and compare their accuracy and training times: (i) a recurrent model using LSTMs; (ii) a fully convolutional model; and (iii) the recently proposed transformer model. The recurrent and fully convolutional models are trained with a Connectionist Temporal Classification loss and use an explicit language model for decoding, the transformer is a sequence-to-sequence model. Our best performing model improves the state-of-the-art word error rate on the challenging BBC-Oxford Lip Reading Sentences 2 (LRS2) benchmark dataset by over 20 percent. As a further contribution we investigate the fully convolutional model when used for online (real time) lip reading of continuous speech, and show that it achieves high performance with low latency.

Many real-world problems, such as network packet routing and urban traffic control, are naturally modeled as multi-agent reinforcement learning (RL) problems. However, existing multi-agent RL methods typically scale poorly in the problem size. Therefore, a key challenge is to translate the success of deep learning on single-agent RL to the multi-agent setting. A major stumbling block is that independent Q-learning, the most popular multi-agent RL method, introduces nonstationarity that makes it incompatible with the experience replay memory on which deep Q-learning relies. This paper proposes two methods that address this problem: 1) using a multi-agent variant of importance sampling to naturally decay obsolete data and 2) conditioning each agent’s value function on a fingerprint that disambiguates the age of the data sampled from the replay memory. Results on a challenging decentralised variant of StarCraft unit micromanagement confirm that these methods enable the successful combination of experience replay with multi-agent RL.