Language is often used to describe physical interaction, yet most 3D human pose estimation methods overlook this rich source of information. We bridge this gap by leveraging large multimodal models (LMMs) as priors for reconstructing contact poses, offering a scalable alternative to traditional methods that rely on human annotations or motion capture data. Our approach extracts contact-relevant descriptors from an LMM and translates them into tractable losses to constrain 3D human pose optimization. Despite its simplicity, our method produces compelling reconstructions for both two-person interactions and self-contact scenarios, accurately capturing the semantics of physical and social interactions. Our results demonstrate that LMMs can serve as powerful tools for contact prediction and pose estimation, offering an alternative to costly manual human annotations or motion capture data. Our code is publicly available at https://prosepose.github.io.

We introduce a simple framework for predicting the behavior of an agent in multi-agent settings. In contrast to autoregressive (AR) tasks, such as language processing, our focus is on scenarios with multiple agents whose interactions are shaped by physical constraints and internal motivations. To this end, we propose Poly-Autoregressive (PAR) modeling, which forecasts an ego agent’s future behavior by reasoning about the ego agent’s state history and the past and current states of other interacting agents. At its core, PAR represents the behavior of all agents as a sequence of tokens, each representing an agent’s state at a specific timestep. With minimal data pre-processing changes, we show that PAR can be applied to three different problems: human action forecasting in social situations, trajectory prediction for autonomous vehicles, and object pose forecasting during hand-object interaction. Using a small proof-of-concept transformer backbone, PAR outperforms AR across these three scenarios.

This paper investigates visual analogical reasoning in large multimodal models (LMMs) compared to human adults and children. A “visual analogy” is an abstract rule inferred from one image and applied to another. While benchmarks exist for testing visual reasoning in LMMs, they require advanced skills and omit basic visual analogies that even young children can make. Inspired by developmental psychology, we propose a new benchmark of 4,300 visual transformations of everyday objects to test LMMs on visual analogical reasoning and compare them to children (ages three to five) and to adults. We structure the evaluation into three stages: identifying what changed (e.g., color, number, etc.), how it changed (e.g., added one object), and applying the rule to new scenarios. Our findings show that while GPT-o1, GPT-4V, LLaVA-1.5, and MANTIS identify the “what” effectively, they struggle with quantifying the “how” and extrapolating this rule to new objects. In contrast, children and adults exhibit much stronger analogical reasoning at all three stages. Additionally, the strongest tested model, GPT-o1, performs better in tasks involving simple surface-level visual attributes like color and size, correlating with quicker human adult response times. Conversely, more complex tasks such as number, rotation, and reflection, which necessitate extensive cognitive processing and understanding of extrinsic spatial properties in the physical world, present more significant challenges. Altogether, these findings highlight the limitations of training models on data that primarily consists of 2D images and text.

We present a framework for generating appropriate facial responses from a listener in dyadic social interactions based on the speaker’s words. Given an input transcription of the speaker’s words with their timestamps, our approach autoregressively predicts a response of a listener: a sequence of listener facial gestures, quantized using a VQ-VAE. Since gesture is a language component, we propose treating the quantized atomic motion elements as additional language token inputs to a transformer-based large language model. Initializing our transformer with the weights of a language model pre-trained only on text results in significantly higher quality listener responses than training a transformer from scratch. We show that our generated listener motion is fluent and reflective of language semantics through quantitative metrics and a qualitative user study. In our evaluation, we analyze the model’s ability to utilize temporal and semantic aspects of spoken text.

We present a framework for modeling interactional communication in dyadic conversations: given multimodal inputs of a speaker, we autoregressively output multiple possibilities of corresponding listener motion. We combine the motion and speech audio of the speaker using a motion-audio cross attention transformer. Furthermore, we enable non-deterministic prediction by learning a discrete latent representation of realistic listener motion with a novel motion-encoding VQ-VAE. Our method organically captures the multimodal and non-deterministic nature of nonverbal dyadic interactions. Moreover, it produces realistic 3D listener facial motion synchronous with the speaker (see video). We demonstrate that our method outperforms baselines qualitatively and quantitatively via a rich suite of experiments. To facilitate this line of research, we introduce a novel and large in-the-wild dataset of dyadic conversations. Code, data, and videos available at https://evonneng.github.io/learning2listen/

We introduce a non-parametric approach for infinite video texture synthesis using a representation learned via contrastive learning. We take inspiration from Video Textures [43], which showed that plausible new videos could be generated from a single one by stitching its frames together in a novel yet consistent order. This classic work, however, was constrained by its use of hand-designed distance metrics, limiting its use to simple, repetitive videos. We draw on recent techniques from self-supervised learning to learn this distance metric, allowing us to compare frames in a manner that scales to more challenging dynamics, and to condition on other data, such as audio. We learn representations for video frames and frame-to-frame transition probabilities by fitting a video-specific model trained using contrastive learning. To synthesize a texture, we randomly sample frames with high transition probabilities to generate diverse temporally smooth videos with novel sequences and transitions. The model naturally extends to an audio-conditioned setting without requiring any finetuning. Our model outperforms baselines on human perceptual scores, can handle a diverse range of input videos, and can combine semantic and audiovisual cues in order to synthesize videos that synchronize well with an audio signal.

We propose a novel learned deep prior of body motion for 3D hand shape synthesis and estimation in the domain of conversational gestures. Our model builds upon the insight that body motion and hand gestures are strongly correlated in non-verbal communication settings. We formulate the learning of this prior as a prediction task of 3D hand shape over time given body motion input alone. Trained with 3D pose estimations obtained from a large-scale dataset of internet videos, our hand prediction model produces convincing 3D hand gestures given only the 3D motion of the speaker’s arms as input. We demonstrate the efficacy of our method on hand gesture synthesis from body motion input, and as a strong body prior for single-view image-based 3D hand pose estimation. We demonstrate that our method outperforms previous state-of-the-art approaches and can generalize beyond the monologue-based training data to multiperson conversations.

Human speech is often accompanied by hand and arm gestures. We present a method for cross-modal translation from "in-the-wild" monologue speech of a single speaker to their conversational gesture motion. We train on unlabeled videos for which we only have noisy pseudo ground truth from an automatic pose detection system. Our proposed model significantly outperforms baseline methods in a quantitative comparison. To support research toward obtaining a computational understanding of the relationship between gesture and speech, we release a large video dataset of person-specific gestures.

This paper presents a simple method for “do as I do” motion transfer: given a source video of a person dancing, we can transfer that performance to a novel (amateur) target after only a few minutes of the target subject performing standard moves. We approach this problem as video-to-video translation using pose as an intermediate representation. To transfer the motion, we extract poses from the source subject and apply the learned pose-to-appearance mapping to generate the target subject. We predict two consecutive frames for temporally coherent video results and introduce a separate pipeline for realistic face synthesis. Although our method is quite simple, it produces surprisingly compelling results (see video). This motivates us to also provide a forensics tool for reliable synthetic content detection, which is able to distinguish videos synthesized by our system from real data. In addition, we release a first-of-its-kind open-source dataset of videos that can be legally used for training and motion transfer.

Multi-stage code examples present multiple versions of a program where each stage increases the overall complexity of the code. In order to acquire strategies of program construction using a new language or API, programmers consult multi-stage code examples in books, tutorials and online videos. Authoring multi-stage code examples is currently a tedious process, as it involves keeping several stages of code synchronized in the face of edits and error corrections. We document these difficulties with a formative study examining how programmers author multi-stage code examples. We then present an IDE extension that helps authors create multi-stage code examples by propagating changes (insertions, deletions and modifications) to multiple saved versions of their code. Our system adapts revision control algorithms to the specific task of evolving example code. An informal evaluation finds that taking snapshots of a program as it is being developed and editing these snapshots in hindsight help users in creating multi-stage code examples.

We present a system that lets analysts use paid crowd workers to explore data sets and helps analysts interactively examine and build upon workers' insights. We take advantage of the fact that, for many types of data, independent crowd workers can readily perform basic analysis tasks like examining views and generating explanations for trends and patterns. However, workers operating in parallel can often generate redundant explanations. Moreover, because workers have different competencies and domain knowledge, some responses are likely to be more plausible than others. To efficiently utilize the crowd's work, analysts must be able to quickly identify and consolidate redundant responses and determine which explanations are the most plausible. In this paper, we demonstrate several crowd-assisted techniques to help analysts make better use of crowdsourced explanations: (1) We explore crowd-assisted strategies that utilize multiple workers to detect redundant explanations. We introduce color clustering with representative selection--a strategy in which multiple workers cluster explanations and we automatically select the most-representative result--and show that it generates clusterings that are as good as those produced by experts. (2) We capture explanation provenance by introducing highlighting tasks and capturing workers' browsing behavior via an embedded web browser, and refine that provenance information via source-review tasks. We expose this information in an explanation-management interface that allows analysts to interactively filter and sort responses, select the most plausible explanations, and decide which to explore further.

Keyword-based image search engines are hindered by the lack of proper labeling for images in their indices. In many cases the labels do not agree with the contents of the image itself, since images are generally indexed by their filename and the surrounding text in the Web page. Another popular approach to image search, content based image retrieval, suffers from a gap between the available low level data and the semantic needs of user searches. To overcome these problems we suggest human annotation of images with natural language descriptions. To this end we present Phetch, an engaging multiplayer game that allows people to attach accurate explanatory text captions to arbitrary images on the Web. People play the game because it is fun, and as a side effect we collect valuable information that can be applied towards improving image retrieval. Furthermore, the game can also be used for other novel applications.

Images on the Web present a major accessibility issue for the visually impaired, mainly because the majority of them do not have proper captions. This paper addresses the problem of attaching proper explanatory text descriptions to arbitrary images on the Web. To this end, we introduce Phetch, an enjoyable computer game that collects explanatory descriptions of images. People play the game because it is fun, and as a side effect of game play we collect valuable information. Given any image from the World Wide Web, Phetch can output a correct annotation for it. The collected data can be applied towards significantly improving Web accessibility. In addition to improving accessibility, Phetch is an example of a new class of games that provide entertainment in exchange for human processing power. In essence, we solve a typical computer vision problem with HCI tools alone.