We study the connection between audio observations and the underlying physics of a mundane yet intriguing everyday activity: pouring liquids. Given only the sound of liquid pouring into a container, our objective is to automatically infer physical properties such as the liquid level, the shape and size of the container, the pouring rate and the time to fill. To this end, we: (i) show in theory that these properties can be determined from the fundamental frequency (pitch); (ii) train a pitch detection model with supervision from simulated data and visual data with a physics-inspired objective; (iii) introduce a new large dataset of real pouring videos for a systematic study; (iv) show that the trained model can indeed infer these physical properties for real data; and finally, (v) we demonstrate strong generalization to various container shapes, other datasets, and in-the-wild YouTube videos. Our work presents a keen understanding of a narrow yet rich problem at the intersection of acoustics, physics, and learning. It opens up applications to enhance multisensory perception in robotic pouring.

A fundamental aspect of compositional reasoning in a video is associating people and their actions across time. Recent years have seen great progress in general-purpose vision/video models and a move towards long-video understanding. While exciting, we take a step back and ask: are today’s models good at compositional reasoning on short videos? To this end, we introduce VELOCITI, a benchmark to study Video-LLMs by disentangling and assessing the comprehension of agents, actions, and their associations across multiple events. We adopt the Video-Language Entailment setup and propose StrictVLE that requires correct classification (rather than ranking) of the positive and negative caption. We evaluate several models and observe that even the best, LLaVA-OneVision (44.5%) and Gemini-1.5-Pro (49.3%), are far from human accuracy at 93.0%. Results show that action understanding lags behind agents, and negative captions created using entities appearing in the video perform worse than those obtained from pure text manipulation. We also present challenges with ClassicVLE and multiple-choice (MC) evaluation, strengthening our preference for StrictVLE. Finally, we validate that our benchmark requires visual inputs of multiple frames making it ideal to study video-language compositional reasoning.

Audio descriptions (ADs) narrate important visual details in movies, enabling Blind and Low Vision (BLV) users to understand narratives and appreciate visual details.Existing works in automatic AD generation mostly focus on few-second trimmed clips, and evaluate them by comparing against a single groundtruth reference AD.However, writing ADs is inherently subjective.Through alignment and analysis of two independent AD tracks for the same movies, we quantify the subjectivity in when and whether to describe, and what and how to highlight.Thus, we show that working with trimmed clips is inadequate.We propose ADQA, a QA benchmark that evaluates ADs at the level of few-minute long, coherent video segments, testing whether they would help BLV users understand the story and appreciate visual details.ADQA features visual appreciation (VA) questions about visual facts and narrative understanding (NU) questions based on the plot.Through ADQA, we show that current AD generation methods lag far behind humanauthored ADs.We conclude with several recommendations for future work and introduce a public leaderboard for benchmarking.Generate Question Generate Question Generate Answer . . . . . .Dialog: I wish that for only one day Dad couldn't tell a lie

We study the connection between audio-visual observations and the underlying physics of a mundane yet intriguing everyday activity: pouring liquids. Given only the sound of liquid pouring into a container, our objective is to automatically infer physical properties such as the liquid level, the shape and size of the container, the pouring rate, and the time to fill. To this end, we: (i) show in theory that these properties can be determined from the fundamental frequency (pitch); (ii) train a pitch detection model with supervision from simulated data and visual data with a physics-inspired objective; (iii) introduce a new large dataset of real pouring videos for a systematic study; (iv) show that the trained model can indeed infer these physical properties for real data; and finally, (v) we demonstrate strong generalization to various container shapes, other datasets, and in-the-wild YouTube videos. Our work presents a keen understanding of a narrow yet rich problem at the intersection of acoustics, physics, and learning. It opens up applications to enhance multisensory perception in robotic pouring.

Understanding what makes a video memorable has important applications in advertising and education technology. Towards this goal, we investigate spatiotemporal attention mechanisms underlying video memorability. Different from previous works that fuse multiple features, we adopt a simple CNN+ Transformer architecture that enables analysis of spatio-temporal attention while matching state-of-the-art (SoTA) performance on video memorability prediction. We compare model attention against human gaze fixations collected through a small-scale eye-tracking study where humans perform the video memory task. We uncover the following insights: (i) Quantitative saliency metrics show that our model, trained only to predict a memorability score, exhibits similar spatial attention patterns to human gaze, especially for more memorable videos. (ii) The model assigns greater importance to initial frames in a video, mimicking human attention patterns. (iii) Panoptic segmentation reveals that both (model and humans) assign a greater share of attention to things and less attention to stuff as compared to their occurrence probability.

Recent advances in text-to-image models have enabled a new era of creative and controllable image generation. However, generating compositional scenes with multiple subjects and attributes remains a significant challenge. To enhance user control over subject placement, several layout-guided methods have been proposed. However, these methods face numerous challenges, particularly in compositional scenes. Unintended subjects often appear outside the layouts, generated images can be out-of-distribution and contain unnatural artifacts, or attributes bleed across subjects, leading to incorrect visual outputs. In this work, we propose MALeR, a method that addresses each of these challenges. Given a text prompt and corresponding layouts, our method prevents subjects from appearing outside the given layouts while being in-distribution. Additionally, we propose a masked, attribute-aware binding mechanism that prevents attribute leakage, enabling accurate rendering of subjects with multiple attributes, even in complex compositional scenes. Qualitative and quantitative evaluation demonstrates that our method achieves superior performance in compositional accuracy, generation consistency, and attribute binding compared to previous work. MALeR is particularly adept at generating images of scenes with multiple subjects and multiple attributes per subject.

Characters are an important aspect of any storyline and identifying and including them in descriptions is necessary for story understanding. While previous work has largely ignored identity and generated captions with some-one (anonymized names), recent work formulates id-aware captioning as a fill-in-the-blanks (FITB) task, where, given a caption with blanks, the goal is to predict person id labels. However, to predict captions with ids, a two-stage approach is required: first predict captions with someone, then fill in identities. In this work, we present a new single stage approach that can seamlessly switch between id-aware caption generation or FITB when given a caption with blanks. Our model, Movie-Identity Captioner (MI-Cap), uses a shared auto-regressive decoder that benefits from training with FITB and full-caption generation ob-jectives, while the encoder can benefit from or disregard captions with blanks as input. Another challenge with id-aware captioning is the lack of a metric to capture subtle differences between person ids. To this end, we introduce iSPICE, a caption evaluation metric that focuses on identity tuples created through intermediate scene graphs. We eval-uate MICap on Large-Scale Movie Description Challenge (LSMDC), where we show a 4.2% improvement in FITB accuracy, and a 1-2% bump in classic captioning metrics.

We introduce multimodal story summarization by lever-aging TV episode recaps - short video sequences interweaving key story moments from previous episodes to bring viewers up to speed. We propose PlotSnap, a dataset featuring two crime thriller TV shows with rich recaps and long episodes of 40 minutes. Story summarization labels are unlocked by matching recap shots to corresponding sub-stories in the episode. We propose a hierarchical model TaleSumm that processes entire episodes by creating compact shot and dialog representations, and predicts importance scores for each video shot and dialog utterance by enabling interactions between local story groups. Unlike traditional summarization, our method extracts multiple plot points from long videos. We present a thorough evaluation on story summarization, including promising cross-series generalization. TaleSumm also shows good results on classic video summarization benchmarks.

The limited generalization of coreference resolution (CR) models has been a major bottleneck in the task's broad application.Prior work has identified annotation differences, especially for mention detection, as one of the main reasons for the generalization gap and proposed using additional annotated target domain data.Rather than relying on this additional annotation, we propose an alternative referential task, Major Entity Identification (MEI), where we: (a) assume the target entities to be specified in the input, and (b) limit the task to only the frequent entities.Through extensive experiments, we demonstrate that MEI models generalize well across domains on multiple datasets with supervised models and LLM-based few-shot prompting.Additionally, MEI fits the classification framework, which enables the use of robust and intuitive classification-based metrics.Finally, MEI is also of practical use as it allows a user to search for all mentions of a particular entity or a group of entities of interest.

Modelling and understanding time remains a challenge in contemporary video understanding models. With language emerging as a key driver towards powerful generalization, it is imperative for foundational video-language models to have a sense of time. In this paper, we consider a specific aspect of temporal understanding: consistency of time order as elicited by before/after relations. We establish that seven existing video-language models struggle to understand even such simple temporal relations. We then question whether it is feasible to equip these foundational models with temporal awareness without re-training them from scratch. Towards this, we propose a temporal adaptation recipe on top of one such model, VideoCLIp, based on post-pretraining on a small amount of video-text data. We conduct a zero-shot evaluation of the adapted models on six datasets for three downstream tasks which require varying degrees of time awareness. We observe encouraging performance gains especially when the task needs higher time awareness. Our work serves as a first step towards probing and instilling a sense of time in existing video-language models without the need for data and compute-intense training from scratch.

Movie story analysis requires understanding characters' emotions and mental states. Towards this goal, we formulate emotion understanding as predicting a diverse and multi-label set of emotions at the level of a movie scene and for each character. We propose EmoTx, a multimodal Transformer-based architecture that ingests videos, multiple characters, and dialog utterances to make joint predictions. By leveraging annotations from the MovieGraphs dataset [72], we aim to predict classic emotions (e.g. happy, angry) and other mental states (e.g. honest, helpful). We conduct experiments on the most frequently occurring 10 and 25 labels, and a mapping that clusters 181 labels to 26. Ablation studies and comparison against adapted state-of-the-art emotion recognition approaches shows the effectiveness of EmoTx. Analyzing EmoTx's self-attention scores reveals that expressive emotions often look at character tokens while other mental states rely on video and dialog cues.

Over the last decade, online lecture videos have become increasingly popular and have experienced a meteoric rise during the pandemic. However, video-language research has primarily focused on instructional videos or movies, and tools to help students navigate the growing online lectures are lacking. Our first contribution is to facilitate research in the educational domain by introducing AVLectures, a large-scale dataset consisting of 86 courses with over 2,350 lectures covering various STEM subjects. Each course contains video lectures, transcripts, OCR outputs for lecture frames, and optionally lecture notes, slides, assignments, and related educational content that can inspire a variety of tasks. Our second contribution is introducing video lecture segmentation that splits lectures into bite-sized topics. Lecture clip representations leverage visual, textual, and OCR cues and are trained on a pretext self-supervised task of matching the narration with the temporally aligned visual content. We formulate lecture segmentation as an unsupervised task and use these representations to generate segments using a temporally consistent 1-nearest neighbor algorithm, TW-FINCH [44]. We evaluate our method on 15 courses and compare it against various visual and textual baselines, outperforming all of them. Our comprehensive ablation studies also identify the key factors driving the success of our approach.

Following language instructions to navigate in unseenaenvironments is a challenging problem for autonomous embodied agents. The agent not only needs to ground languages in visual scenes, but also should explore the environment to reach its target. In this work, we propose a dual-scale graph transformer (DUET) for joint long-term action planning and fine-grained cross-modal understanding. We build a topological map on-the-fly to enable efficient exploration in global action space. To balance the complexity of large action space reasoning and fine-grained language grounding, we dynamically combine a fine-scale encoding over local observations and a coarse-scale encoding on a global map via graph transformers. The proposed approach, DUET, significantly outperforms state-of-the-art methods on goal-oriented vision-and-language navigation (VLN) benchmarks REVERIE and SOON. It also improves the success rate on the fine-grained VLN benchmark R2R.

We aim to teach robots to perform simple object manipulation tasks by watching a single video demonstration. Towards this goal, we propose an optimization approach that outputs a coarse and temporally evolving 3D scene to mimic the action demonstrated in the input video. Similar to previous work, a differentiable renderer ensures perceptual fidelity between the 3D scene and the 2D video. Our key novelty lies in the inclusion of a differentiable approach to solve a set of Ordinary Differential Equations (ODEs) that allows us to approximately model laws of physics such as gravity, friction, and hand-object or object-object interactions. This not only enables us to dramatically improve the quality of estimated hand and object states, but also produces physically admissible trajectories that can be directly translated to a robot without the need for costly reinforcement learning. We evaluate our approach on a 3D reconstruction task that consists of 54 video demonstrations sourced from 9 actions such as pull something from right to left or put something in front of something. Our approach improves over previous state-of-the-art by almost 30%, demonstrating superior quality on especially challenging actions involving physical interactions of two objects such as put something onto something. Finally, we showcase the learned skills on a Franka Emika Panda robot.

Localizing objects in 3D scenes based on natural language requires understanding and reasoning about spatial relations. In particular, it is often crucial to distinguish similar objects referred by the text, such as "the left most chair" and "a chair next to the window". In this work we propose a language-conditioned transformer model for grounding 3D objects and their spatial relations. To this end, we design a spatial self-attention layer that accounts for relative distances and orientations between objects in input 3D point clouds. Training such a layer with visual and language inputs enables to disambiguate spatial relations and to localize objects referred by the text. To facilitate the cross-modal learning of relations, we further propose a teacher-student approach where the teacher model is first trained using ground-truth object labels, and then helps to train a student model using point cloud inputs. We perform ablation studies showing advantages of our approach. We also demonstrate our model to significantly outperform the state of the art on the challenging Nr3D, Sr3D and ScanRefer 3D object grounding datasets.

Dense video understanding requires answering several questions such as who is doing what to whom, with what, how, why, and where. Recently, Video Situation Recognition (VidSitu) is framed as a task for structured prediction of multiple events, their relationships, and actions and various verb-role pairs attached to descriptive entities. This task poses several challenges in identifying, disambiguating, and co-referencing entities across multiple verb-role pairs, but also faces some challenges of evaluation. In this work, we propose the addition of spatio-temporal grounding as an essential component of the structured prediction task in a weakly supervised setting, and present a novel three stage Transformer model, VideoWhisperer, that is empowered to make joint predictions. In stage one, we learn contextualised embeddings for video features in parallel with key objects that appear in the video clips to enable fine-grained spatio-temporal reasoning. The second stage sees verb-role queries attend and pool information from object embeddings, localising answers to questions posed about the action. The final stage generates these answers as captions to describe each verb-role pair present in the video. Our model operates on a group of events (clips) simultaneously and predicts verbs, verb-role pairs, their nouns, and their grounding on-the-fly. When evaluated on a grounding-augmented version of the VidSitu dataset, we observe a large improvement in entity captioning accuracy, as well as the ability to localize verb-roles without grounding annotations at training time.

Vision-and-language navigation (VLN) aims to enable embodied agents to navigate in realistic environments using natural language instructions. Given the scarcity of domain-specific training data and the high diversity of image and language inputs, the generalization of VLN agents to unseen environments remains challenging. Recent methods explore pretraining to improve generalization, however, the use of generic image-caption datasets or existing small-scale VLN environments is suboptimal and results in limited improvements. In this work, we introduce BnB1, a large-scale and diverse in-domain VLN dataset. We first collect image-caption (IC) pairs from hundreds of thousands of listings from online rental marketplaces. Using IC pairs we next propose automatic strategies to generate millions of VLN path-instruction (PI) pairs. We further propose a shuffling loss that improves the learning of temporal order inside PI pairs. We use BnB to pretrain our Airbert2model that can be adapted to discriminative and generative settings and show that it outperforms state of the art for Room-to-Room (R2R) navigation and Remote Referring Expression (REVERIE) benchmarks. Moreover, our in-domain pretraining significantly increases performance on a challenging few-shot VLN evaluation, where we train the model only on VLN instructions from a few houses.

Interactions between people are often governed by their relationships. On the flip side, social relationships are built upon several interactions. Two strangers are more likely to greet and introduce themselves while becoming friends over time. We are fascinated by this interplay between interactions and relationships, and believe that it is an important aspect of understanding social situations. In this work, we propose neural models to learn and jointly predict interactions, relationships, and the pair of characters that are involved. We note that interactions are informed by a mixture of visual and dialog cues, and present a multimodal architecture to extract meaningful information from them. Localizing the pair of interacting characters in video is a time-consuming process, instead, we train our model to learn from clip-level weak labels. We evaluate our models on the MovieGraphs dataset and show the impact of modalities, use of longer temporal context for predicting relationships, and achieve encouraging performance using weak labels as compared with ground-truth labels. Code is online.

A good clustering algorithm can discover natural groupings in data. These groupings, if used wisely, provide a form of weak supervision for learning representations. In this work, we present Clustering-based Contrastive Learning (CCL), a new clustering-based representation learning approach that uses labels obtained from clustering along with video constraints to learn discriminative face features. We demonstrate our method on the challenging task of learning representations for video face clustering. Through several ablation studies, we analyze the impact of creating pair-wise positive and negative labels from different sources. Experiments on three challenging video face clustering datasets: BBT-0101, BF-0502, and ACCIO show that CCL achieves a new state-of-the-art on all datasets.

Analyzing the story behind TV series and movies often requires understanding who the characters are and what they are doing. With improving deep face models, this may seem like a solved problem. However, as face detectors get better, clustering/identification needs to be revisited to address increasing diversity in facial appearance. In this paper, we address video face clustering using unsupervised methods. Our emphasis is on distilling the essential information, identity, from the representations obtained using deep pre-trained face networks. We propose a self-supervised Siamese network that can be trained without the need for video/track based supervision, and thus can also be applied to image collections. We evaluate our proposed method on three video face clustering datasets. The experiments show that our methods outperform current state-of-the-art methods on all datasets. Video face clustering is lacking a common benchmark as current works are often evaluated with different metrics and/or different sets of face tracks. The datasets and code are available at https://github.com/vivoutlaw/SSIAM.

Learning text-video embeddings usually requires a dataset of video clips with manually provided captions. However, such datasets are expensive and time consuming to create and therefore difficult to obtain on a large scale. In this work, we propose instead to learn such embeddings from video data with readily available natural language annotations in the form of automatically transcribed narrations. The contributions of this work are three-fold. First, we introduce HowTo100M: a large-scale dataset of 136 million video clips sourced from 1.22M narrated instructional web videos depicting humans performing and describing over 23k different visual tasks. Our data collection procedure is fast, scalable and does not require any additional manual annotation. Second, we demonstrate that a text-video embedding trained on this data leads to state-of-the-art results for text-to-video retrieval and action localization on instructional video datasets such as YouCook2 or CrossTask. Finally, we show that this embedding transfers well to other domains: fine-tuning on generic Youtube videos (MSR-VTT dataset) and movies (LSMDC dataset) outperforms models trained on these datasets alone. Our dataset, code and models are publicly available.

Understanding videos such as TV series and movies requires analyzing who the characters are and what they are doing. We address the challenging problem of clustering face tracks based on their identity. Different from previous work in this area, we choose to operate in a realistic and difficult setting where: (i) the number of characters is not known a priori; and (ii) face tracks belonging to minor or background characters are not discarded. To this end, we propose Ball Cluster Learning (BCL), a supervised approach to carve the embedding space into balls of equal size, one for each cluster. The learned ball radius is easily translated to a stopping criterion for iterative merging algorithms. This gives BCL the ability to estimate the number of clusters as well as their assignment, achieving promising results on commonly used datasets. We also present a thorough discussion of how existing metric learning literature can be adapted for this task.

There is growing interest in artificial intelligence to build socially intelligent robots. This requires machines to have the ability to "read" people's emotions, motivations, and other factors that affect behavior. Towards this goal, we introduce a novel dataset called MovieGraphs which provides detailed, graph-based annotations of social situations depicted in movie clips. Each graph consists of several types of nodes, to capture who is present in the clip, their emotional and physical attributes, their relationships (i.e., parent/child), and the interactions between them. Most interactions are associated with topics that provide additional details, and reasons that give motivations for actions. In addition, most interactions and many attributes are grounded in the video with time stamps. We provide a thorough analysis of our dataset, showing interesting common-sense correlations between different social aspects of scenes, as well as across scenes over time. We propose a method for querying videos and text with graphs, and show that: 1) our graphs contain rich and sufficient information to summarize and localize each scene; and 2) subgraphs allow us to describe situations at an abstract level and retrieve multiple semantically relevant situations. We also propose methods for interaction understanding via ordering, and reason understanding. MovieGraphs is the first benchmark to focus on inferred properties of human-centric situations, and opens up an exciting avenue towards socially-intelligent AI agents.

We are interested in enabling automatic 4D cinema by parsing physical and special effects from untrimmed movies. These include effects such as physical interactions, water splashing, light, and shaking, and are grounded to either a character in the scene or the camera. We collect a new dataset referred to as the Movie4D dataset which annotates over 9K effects in 63 movies. We propose a Conditional Random Field model atop a neural network that brings together visual and audio information, as well as semantics in the form of person tracks. Our model further exploits correlations of effects between different characters in the clip as well as across movie threads. We propose effect detection and classification as two tasks, and present results along with ablation studies on our dataset, paving the way towards 4D cinema in everyone's homes.

Collecting training images for all visual categories is not only expensive but also impractical. Zero-shot learning (ZSL), especially using attributes, offers a pragmatic solution to this problem. However, at test time most attribute-based methods require a full description of attribute associations for each unseen class. Providing these associations is time consuming and often requires domain specific knowledge. In this work, we aim to carry out attribute-based zero-shot classification in an unsupervised manner. We propose an approach to learn relations that couples class embeddings with their corresponding attributes. Given only the name of an unseen class, the learned relationship model is used to automatically predict the class-attribute associations. Furthermore, our model facilitates transferring attributes across data sets without additional effort. Integrating knowledge from multiple sources results in a significant additional improvement in performance. We evaluate on two public data sets: Animals with Attributes and aPascal/aYahoo. Our approach outperforms state-of the-art methods in both predicting class-attribute associations and unsupervised ZSL by a large margin.

We introduce the MovieQA dataset which aims to evaluate automatic story comprehension from both video and text. The dataset consists of 14,944 questions about 408 movies with high semantic diversity. The questions range from simpler "Who" did "What" to "Whom", to "Why" and "How" certain events occurred. Each question comes with a set of five possible answers, a correct one and four deceiving answers provided by human annotators. Our dataset is unique in that it contains multiple sources of information – video clips, plots, subtitles, scripts, and DVS [32]. We analyze our data through various statistics and methods. We further extend existing QA techniques to show that question-answering with such open-ended semantics is hard. We make this data set public along with an evaluation benchmark to encourage inspiring work in this challenging domain.

Person identification in TV series has been a popular research topic over the last decade. In this area, most approaches either use manually annotated data or extract character supervision from a combination of subtitles and transcripts. However, both approaches have key drawbacks that hinder application of these methods at a large scale - manual annotation is expensive and transcripts are often hard to obtain. We investigate the topic of automatically labeling all character appearances in TV series using information obtained solely from subtitles. This task is extremely difficult as the dialogs between characters provide very sparse and weakly supervised data. We address these challenges by exploiting recent advances in face descriptors and Multiple Instance Learning methods. We propose methods to create MIL bags and evaluate and discuss several MIL techniques. The best combination achieves an average precision over 80% on three diverse TV series. We demonstrate that only using subtitles provides good results on identifying characters in TV series and wish to encourage the community towards this problem.

Film adaptations of novels often visually display in a few shots what is described in many pages of the source novel. In this paper we present a new problem: to align book chapters with video scenes. Such an alignment facilitates finding differences between the adaptation and the original source, and also acts as a basis for deriving rich descriptions from the novel for the video clips. We propose an efficient method to compute an alignment between book chapters and video scenes using matching dialogs and character identities as cues. A major consideration is to allow the alignment to be non-sequential. Our suggested shortest path based approach deals with the non-sequential alignments and can be used to determine whether a video scene was part of the original book. We create a new data set involving two popular novel-to-film adaptations with widely varying properties and compare our method against other text-to-video alignment baselines. Using the alignment, we present a qualitative analysis of describing the video through rich narratives obtained from the novel.

Video face recognition is a very popular task and has come a long way. The primary challenges such as illumination, resolution and pose are well studied through multiple data sets. However there are no video-based data sets dedicated to study the effects of aging on facial appearance. We present a challenging face track data set, Harry Potter Movies Aging Data set (Accio1), to study and develop age invariant face recognition methods for videos. Our data set not only has strong challenges of pose, illumination and distractors, but also spans a period of ten years providing substantial variation in facial appearance. We propose two primary tasks: within and across movie face track retrieval; and two protocols which differ in their freedom to use external data. We present baseline results for the retrieval performance using a state-of-the-art face track descriptor. Our experiments show clear trends of reduction in performance as the age gap between the query and database increases. We will make the data set publicly available for further exploration in age-invariant video face recognition.

We present a novel method for comparing tracks by means of a time pooled track kernel. In contrast to spatial or feature-space pooling, the track kernel pools base kernel results within tracks over time. It includes as special cases frame-wise classification on the one hand and the normalized sum kernel on the other hand. We also investigate non-Mercer instantiations of the track kernel and obtain good results despite its Gram matrices not being positive semidefinite. Second, the track kernel matrices in general require less memory than single frame kernels, allowing to process larger datasets without resorting to subsampling. Finally, the track kernel formulation allows for very fast testing compared to frame-wise classification which is important in settings where user feedback is obtained and quick iterations of re-training and re-testing are required. We apply our approach to the task of video-based person identification in large scale settings and obtain state-of-the art results.

We present a novel way to automatically summarize and represent the storyline of a TV episode by visualizing character interactions as a chart. We also propose a scene detection method that lends itself well to generate over-segmented scenes which is used to partition the video. The positioning of character lines in the chart is formulated as an optimization problem which trades between the aesthetics and functionality of the chart. Using automatic person identification, we present StoryGraphs for 3 diverse TV series encompassing a total of 22 episodes. We define quantitative criteria to evaluate StoryGraphs and also compare them against episode summaries to evaluate their ability to provide an overview of the episode.

Automatic person identification in TV series has gained popularity over the years. While most of the works rely on using face-based recognition, errors during tracking such as false positive face tracks are typically ignored. We propose a variety of methods to remove false positive face tracks and categorize the methods into confidence- and context-based. We evaluate our methods on a large TV series data set and show that up to 75% of the false positive face tracks are removed at the cost of 3.6% true positive tracks. We further show that the proposed method is general and applicable to other detectors or trackers.

We present a novel approach to search for plots in the storyline of structured videos such as TV series. To this end, we propose to align natural language descriptions of the videos, such as plot synopses, with the corresponding shots in the video. Guided by subtitles and person identities the alignment problem is formulated as an optimization task over all possible assignments and solved efficiently using dynamic programming. We evaluate our approach on a novel dataset comprising of the complete season 5 of Buffy the Vampire Slayer, and show good alignment performance and the ability to retrieve plots in the storyline.

We address the problem of person identification in TV series. We propose a unified learning framework for multi-class classification which incorporates labeled and unlabeled data, and constraints between pairs of features in the training. We apply the framework to train multinomial logistic regression classifiers for multi-class face recognition. The method is completely automatic, as the labeled data is obtained by tagging speaking faces using subtitles and fan transcripts of the videos. We demonstrate our approach on six episodes each of two diverse TV series and achieve state-of-the-art performance.

We use contextual constraints for person retrieval in camera networks. We start by formulating a set of general positive and negative constraints on the identities of person tracks in camera networks, such as a person cannot appear twice in the same frame. We then show how these constraints can be used to improve person retrieval. First, we use the constraints to obtain training data in an unsupervised way to learn a general metric that is better suited to discriminate between different people than the Euclidean distance. Second, starting from an initial query track, we enhance the query-set using the constraints to obtain additional positive and negative samples for the query. Third, we formulate the person retrieval task as an energy minimization problem, integrate track scores and constraints in a common framework and jointly optimize the retrieval over all interconnected tracks. We evaluate our approach on the CAVIAR dataset and achieve 22% relative performance improvement in terms of mean average precision over standard retrieval where each track is treated independently.

We describe a probabilistic method for identifying characters in TV series or movies. We aim at labeling every character appearance, and not only those where a face can be detected. Consequently, our basic unit of appearance is a person track (as opposed to a face track). We model each TV series episode as a Markov Random Field, integrating face recognition, clothing appearance, speaker recognition and contextual constraints in a probabilistic manner. The identification task is then formulated as an energy minimization problem. In order to identify tracks without faces, we learn clothing models by adapting available face recognition results. Within a scene, as indicated by prior analysis of the temporal structure of the TV series, clothing features are combined by agglomerative clustering. We evaluate our approach on the first 6 episodes of The Big Bang Theory and achieve an absolute improvement of 20% for person identification and 12% for face recognition.

Mobile phones are playing an important role in changing the socio-economic landscapes of emerging economies like India. A proper voice-based user authentication will help in many new mobile based applications including mobile-commerce and banking. We present our exploration and evaluation of an experimental set-up for user authentication in remote Indian villages using mobile phones and user-selected multilingual spoken passwords. We also present an effective speaker recognition method using a set of novel features called Compressed Feature Dynamics (CFD) which capture the speaker-identity effectively from the speech dynamics contained in the spoken passwords. Early trials demonstrate the effectiveness of the proposed method in handling noisy cell-phone speech. Compared to conventional text-dependent speaker recognition methods, the proposed CFD method delivers competitive performance while significantly reducing storage and computational complexity - an advantage highly beneficial for cell-phone based deployment of such user authentication systems.