Skip to content
DraconiticResearch
Search Sources API
194,230 studies31 sourcesUpdated 21 min ago
Peer-reviewedCC BY 4.0PLOS One

TAL-CGBR: A class-guided boundary regression approach for Temporal Action Localization

  • Jinling Wei1
  • Haolong Chai2
  • Wenwu Hou3
  • Hongwen Yuan
  • Xiao Huang
  • Sufen Huang
  1. 1School of Computer and Computing Science, Hangzhou City University, Hangzhou, China
  2. 2College of Computer Science and Technology, Zhejiang University, Hangzhou, China
  3. 3Department of Physical Education and Aesthetic Education, Hangzhou City University, Hangzhou, China
Open at sourcePDFDOICite

Abstract

Temporal Action Localization (TAL) aims to identify all action categories along with their precise temporal boundaries in untrimmed videos. It serves as a fundamental task supporting various downstream applications, including intelligent video editing, automatic summarization, anomaly detection, and human–computer interaction. Despite significant progress, TAL in real-world scenarios remains challenging due to two critical factors: substantial intra-class variations in action duration, ranging from a few seconds to several minutes, and severe degradation of boundary localization accuracy caused by variations in illumination, background clutter, and camera viewpoints. To address these issues, this paper presents a Temporal Action Localization with Class-Guided Boundary Regression (TAL-CGBR) framework, which incorporates semantic information from the classification branch into the regression network to enhance boundary precision. In addition, a window-based attention mechanism and a temporal hierarchical feature structure are introduced to improve computational efficiency while maintaining high localization accuracy. Experimental evaluations on benchmark datasets demonstrate that the proposed TAL-CGBR achieves competitive performance, attaining 68.5% average mAP on THUMOS14 and 36.7% average mAP on ActivityNet v1.3, ranking second among recent methods and demonstrating a favorable balance between accuracy and computational efficiency.

Related

  1. Revisiting genome-guided RNA localization in bacteria
    Roy Jacobson, Zohar Persky, Yogev Adler +1
    Nature CommunicationsPeer-reviewedCC BY-NC-ND 4.0Full textGenetics & genomicsCrossref
  2. Circuit plasticity coordinates action selection and temporal compression during operant conditioning in Aplysia
    Yuto Momohara, Curtis L. Neveu, John H. Byrne
    Communications BiologyPeer-reviewedCC BY-NC-ND 4.0NeuroscienceNatureCrossref
  3. Morphology-guided dual-level detection and depth localization of cotton terminal buds for precision topping
    Chaoqun Wang, Haojun Wen, Zhaohui Zheng +3
    Frontiers in Plant SciencePeer-reviewedCC BY 4.0Full textPlant biologyCrossrefEurope PMCPMC
  4. Physics-guided explainable machine learning for predicting temporal scour evolution around hybrid spur dikes under unsteady flood hydrographs
    Mehdi Daryaee, Laleh Divband Hafshejani, Seyed Mahmood Kashefipour +1
    Scientific ReportsPeer-reviewedCC BY-NC-ND 4.0Full textEcologyCrossref
  5. Canopy and leaf vein segmentation and topology-guided localization of stem points in tobacco plants using YOLO-ODS
    Jiaxing Qing, Donghao Li, Xiwen Luo +7
    Frontiers in Plant SciencePeer-reviewedCC BY 4.0Full textPlant biologyCrossrefEurope PMC

Cite

BibTeXRIS
BibTeX
@article{wei2026talcgbr,
  title = {{TAL-CGBR: A class-guided boundary regression approach for Temporal Action Localization}},
  author = {Jinling Wei and Haolong Chai and Wenwu Hou and Hongwen Yuan and Xiao Huang and Sufen Huang},
  journal = {PLOS One},
  year = {2026},
  month = oct,
  doi = {10.1371/journal.pone.0360160},
  url = {https://doi.org/10.1371/journal.pone.0360160},
  note = {CC BY 4.0},
}
RIS
TY  - JOUR
TI  - TAL-CGBR: A class-guided boundary regression approach for Temporal Action Localization
AU  - Jinling Wei
AU  - Haolong Chai
AU  - Wenwu Hou
AU  - Hongwen Yuan
AU  - Xiao Huang
AU  - Sufen Huang
PY  - 2026
DA  - 2026/10/12
JO  - PLOS One
DO  - 10.1371/journal.pone.0360160
UR  - https://doi.org/10.1371/journal.pone.0360160
AB  - Temporal Action Localization (TAL) aims to identify all action categories along with their precise temporal boundaries in untrimmed videos. It serves as a fundamental task supporting various downstream applications, including intelligent video editing, automatic summarization, anomaly detection, and human–computer interaction. Despite significant progress, TAL in real-world scenarios remains challenging due to two critical factors: substantial intra-class variations in action duration, ranging from a few seconds to several minutes, and severe degradation of boundary localization accuracy caused by variations in illumination, background clutter, and camera viewpoints. To address these issues, this paper presents a Temporal Action Localization with Class-Guided Boundary Regression (TAL-CGBR) framework, which incorporates semantic information from the classification branch into the regression network to enhance boundary precision. In addition, a window-based attention mechanism and a temporal hierarchical feature structure are introduced to improve computational efficiency while maintaining high localization accuracy. Experimental evaluations on benchmark datasets demonstrate that the proposed TAL-CGBR achieves competitive performance, attaining 68.5% average mAP on THUMOS14 and 36.7% average mAP on ActivityNet v1.3, ranking second among recent methods and demonstrating a favorable balance between accuracy and computational efficiency.
SN  - 1932-6203
ER  -