Do You See What I Am Pointing At? Gesture-Based Egocentric Video Question Answering

2026-03-12

AI Summary

Existing vision-language models struggle to interpret the fine-grained intent conveyed by users’ pointing gestures in first-person videos, primarily due to the absence of dedicated datasets and effective modeling mechanisms. This work presents the first systematic benchmark, EgoPointVQA, comprising both synthetic and real-world egocentric videos annotated for pointing gestures, and introduces Hand Intent Tokens (HINT)—a novel approach that leverages off-the-shelf 3D hand reconstruction models to extract spatiotemporal keypoints, which are then encoded as explicit intent tokens and integrated into the input of multimodal large language models to enhance referential reasoning. The resulting HINT-14B model achieves an average accuracy of 68.1% across six tasks, outperforming InternVL3-14B by 6.6 percentage points. Code, models, and the dataset are publicly released.

Technology Category

• Computer Vision: Large Vision Models
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Agentic search
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Understanding and answering questions based on a user's pointing gesture is essential for next-generation egocentric AI assistants. However, current Multimodal Large Language Models (MLLMs) struggle with such tasks due to the lack of gesture-rich data and their limited ability to infer fine-grained pointing intent from egocentric video. To address this, we introduce EgoPointVQA, a dataset and benchmark for gesture-grounded egocentric question answering, comprising 4000 synthetic and 400 real-world videos across multiple deictic reasoning tasks. Built upon it, we further propose Hand Intent Tokens (HINT), which encodes tokens derived from 3D hand keypoints using an off-the-shelf reconstruction model and interleaves them with the model input to provide explicit spatial and temporal context for interpreting pointing intent. We show that our model outperforms others in different backbones and model sizes. In particular, HINT-14B achieves 68.1% accuracy, on average over 6 tasks, surpassing the state-of-the-art, InternVL3-14B, by 6.6%. To further facilitate the open research, we will release the code, model, and dataset. Project page: https://yuuraa.github.io/papers/choi2026egovqa

Problem

• gesture-based
• egocentric video
• pointing intent
• question answering
• deictic reasoning

Innovation

• Hand Intent Tokens
• Egocentric Video Question Answering
• Pointing Gesture Understanding
• 3D Hand Keypoints
• Deictic Reasoning