Temporal Action Localization

Follow this topic Ask about this topic

Temporal action localization identifies which actions occur in a long video and when each begins and ends, supporting tasks such as video search, sports analysis, and surgical monitoring. Methods use temporal CNNs, transformers, or state-space models to predict action segments, while language-guided grounding and proposal reranking connect video intervals to text queries. Because precise frame-level labels are costly, work also explores weak, point-level, unsupervised, and zero-shot learning, alongside online localization for streaming video. Benchmarks such as THUMOS14 and ActivityNet evaluate detection with mean average precision at different temporal IoU thresholds; a central challenge is balancing accurate boundaries and generalization to new actions against limited supervision and computational cost.

Latest papers

32 papers5 papers in the last four weeks, up 67% on the four weeks before. 0.0% of all new papers.

MGCA-Net: Multi-Grained Category-Aware Network for Open-Vocabulary Temporal Action Localization
Nov 17, 2025Zhenying Fang, Richang HongVideo Action Recognition Temporal Action Detection
School of Computer Science and Information Engineering, Hefei University of Technology, 188065, China
LLM-powered Query Expansion for Enhancing Boundary Prediction in Language-driven Action Localization
May 30, 2025Zirui Shang, Xinxiao Wu, Shuo YangTemporal Action Localization Query Expansion
Beijing Key Laboratory of Intelligent Information Technology, School of Computer Science & Technology, Beijing Institute of Technology, China. · Guangdong Laboratory of Machine Perception and Intelligent Computing, Shenzhen MSU-BIT University, China.

Nov 17, 2025J/K move · Enter open · S save

MGCA-Net: Multi-Grained Category-Aware Network for Open-Vocabulary Temporal Action Localization

Zhenying Fang, Richang Hong

School of Computer Science and Information Engineering, Hefei University of Technology, 188065, China

Video Action Recognition Temporal Action Detection Temporal Video Understanding Temporal Action Localization

Open paper AskPDF arXiv

Open-Vocabulary Temporal Action Localization (OV-TAL) aims to recognize and localize instances of any desired action categories in videos without explicitly curating training data for all categories. Existing methods mostly recognize action categories at a single granularity, which degrades the recognition accuracy of both base and novel action categories. To address these issues, we propose a Multi-Grained Category-Aware Network (MGCA-Net) comprising a localizer, an action presence predictor, a conventional classifier, and a coarse-to-fine classifier. Specifically, the localizer localizes category-agnostic action proposals. For these action proposals, the action presence predictor estimates the probability that they belong to an action instance. At the same time, the conventional classifier predicts the probability of each action proposal over base action categories at the snippet granularity. Novel action categories are recognized by the coarse-to-fine classifier, which first identifies action presence at the video granularity. Finally, it assigns each action proposal to one category from the coarse categories at the proposal granularity. Through coarse-to-fine category awareness for novel actions and the conventional classifier's awareness of base actions, multi-grained category awareness is achieved, effectively enhancing localization performance. Comprehensive evaluations on the THUMOS'14 and ActivityNet-1.3 benchmarks demonstrate that our method achieves state-of-the-art performance. Furthermore, our MGCA-Net achieves state-of-the-art results under the Zero-Shot Temporal Action Localization setting.

Latest papers

MGCA-Net: Multi-Grained Category-Aware Network for Open-Vocabulary Temporal Action Localization

LLM-powered Query Expansion for Enhancing Boundary Prediction in Language-driven Action Localization