Beyond Label Semantics: Language-Guided Action Anatomy for Few-shot Action Recognition

• URL: http://arxiv.org/abs/2507.16287v1
• Date: Tue, 22 Jul 2025 07:16:25 GMT
• Title: Beyond Label Semantics: Language-Guided Action Anatomy for Few-shot Action Recognition
• Authors: Zefeng Qian, Xincheng Yao, Yifei Huang, Chongyang Zhang, Jiangyong Ying, Hong Sun,
• Abstract summary: Few-shot action recognition (FSAR) aims to classify human actions in videos with only a small number of samples labeled per category.<n>We propose Language-Guided Action Anatomy (LGA), a novel framework that goes beyond label semantics.<n>For text, we prompt an off-the-shelf LLM to anatomize labels into sequences of atomic action descriptions.<n>For videos, a Visual Anatomy Module segments actions into atomic video phases to capture the sequential structure of actions.
• Score: 16.07037171149096
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Few-shot action recognition (FSAR) aims to classify human actions in videos with only a small number of labeled samples per category. The scarcity of training data has driven recent efforts to incorporate additional modalities, particularly text. However, the subtle variations in human posture, motion dynamics, and the object interactions that occur during different phases, are critical inherent knowledge of actions that cannot be fully exploited by action labels alone. In this work, we propose Language-Guided Action Anatomy (LGA), a novel framework that goes beyond label semantics by leveraging Large Language Models (LLMs) to dissect the essential representational characteristics hidden beneath action labels. Guided by the prior knowledge encoded in LLM, LGA effectively captures rich spatiotemporal cues in few-shot scenarios. Specifically, for text, we prompt an off-the-shelf LLM to anatomize labels into sequences of atomic action descriptions, focusing on the three core elements of action (subject, motion, object). For videos, a Visual Anatomy Module segments actions into atomic video phases to capture the sequential structure of actions. A fine-grained fusion strategy then integrates textual and visual features at the atomic level, resulting in more generalizable prototypes. Finally, we introduce a Multimodal Matching mechanism, comprising both video-video and video-text matching, to ensure robust few-shot classification. Experimental results demonstrate that LGA achieves state-of-the-art performance across multipe FSAR benchmarks.

Related papers

• Spatio-Temporal Context Prompting for Zero-Shot Action Detection [13.22912547389941]
  We propose a method which can effectively leverage the rich knowledge of visual-language models to perform Person-Context Interaction.<n>To address the challenge of recognizing distinct actions by multiple people at the same timestamp, we design the Interest Token Spotting mechanism.<n>Our method achieves superior results compared to previous approaches and can be further extended to multi-action videos.
  arXiv  Detail & Related papers  (2024-08-28T17:59:05Z)
• TEACH: Temporal Action Composition for 3D Humans [50.97135662063117]
  Given a series of natural language descriptions, our task is to generate 3D human motions that correspond semantically to the text. In particular, our goal is to enable the synthesis of a series of actions, which we refer to as temporal action composition.
  arXiv  Detail & Related papers  (2022-09-09T00:33:40Z)
• Generative Action Description Prompts for Skeleton-based Action Recognition [15.38417530693649]
  We propose a Generative Action-description Prompts (GAP) approach for skeleton-based action recognition. We employ a pre-trained large-scale language model as the knowledge engine to automatically generate text descriptions for body parts movements of actions. Our proposed GAP method achieves noticeable improvements over various baseline models without extra cost at inference.
  arXiv  Detail & Related papers  (2022-08-10T12:55:56Z)
• Modeling Motion with Multi-Modal Features for Text-Based Video Segmentation [56.41614987789537]
  Text-based video segmentation aims to segment the target object in a video based on a describing sentence. We propose a method to fuse and align appearance, motion, and linguistic features to achieve accurate segmentation.
  arXiv  Detail & Related papers  (2022-04-06T02:42:33Z)
• Bridge-Prompt: Towards Ordinal Action Understanding in Instructional Videos [92.18898962396042]
  We propose a prompt-based framework, Bridge-Prompt, to model the semantics across adjacent actions. We reformulate the individual action labels as integrated text prompts for supervision, which bridge the gap between individual action semantics. Br-Prompt achieves state-of-the-art on multiple benchmarks.
  arXiv  Detail & Related papers  (2022-03-26T15:52:27Z)
• Towards Tokenized Human Dynamics Representation [41.75534387530019]
  We study how to segment and cluster videos into recurring temporal patterns in a self-supervised way. We evaluate the frame-wise representation learning step by Kendall's Tau and the lexicon building step by normalized mutual information and language entropy. On the AIST++ and PKU-MMD datasets, actons bring significant performance improvements compared to several baselines.
  arXiv  Detail & Related papers  (2021-11-22T18:59:58Z)
• Learning to Recognize Actions on Objects in Egocentric Video with Attention Dictionaries [51.48859591280838]
  We present EgoACO, a deep neural architecture for video action recognition. It learns to pool action-context-object descriptors from frame level features. Cap uses self-attention with a dictionary of learnable weights to pool from the most relevant feature regions.
  arXiv  Detail & Related papers  (2021-02-16T10:26:04Z)
• Intra- and Inter-Action Understanding via Temporal Action Parsing [118.32912239230272]
  We construct a new dataset developed on sport videos with manual annotations of sub-actions, and conduct a study on temporal action parsing on top. Our study shows that a sport activity usually consists of multiple sub-actions and that the awareness of such temporal structures is beneficial to action recognition. We also investigate a number of temporal parsing methods, and thereon devise an improved method that is capable of mining sub-actions from training data without knowing the labels of them.
  arXiv  Detail & Related papers  (2020-05-20T17:45:18Z)
• FineGym: A Hierarchical Video Dataset for Fine-grained Action Understanding [118.32912239230272]
  FineGym is a new action recognition dataset built on top of gymnastic videos. It provides temporal annotations at both action and sub-action levels with a three-level semantic hierarchy. This new level of granularity presents significant challenges for action recognition.
  arXiv  Detail & Related papers  (2020-04-14T17:55:21Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.