Action-Agnostic Point-Level Supervision for Temporal Action Detection

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yoshida, Shuhei M., Shibata, Takashi, Terao, Makoto, Okatani, Takayuki, Sugiyama, Masashi
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909444675207168
author Yoshida, Shuhei M.
Shibata, Takashi
Terao, Makoto
Okatani, Takayuki
Sugiyama, Masashi
author_facet Yoshida, Shuhei M.
Shibata, Takashi
Terao, Makoto
Okatani, Takayuki
Sugiyama, Masashi
contents We propose action-agnostic point-level (AAPL) supervision for temporal action detection to achieve accurate action instance detection with a lightly annotated dataset. In the proposed scheme, a small portion of video frames is sampled in an unsupervised manner and presented to human annotators, who then label the frames with action categories. Unlike point-level supervision, which requires annotators to search for every action instance in an untrimmed video, frames to annotate are selected without human intervention in AAPL supervision. We also propose a detection model and learning method to effectively utilize the AAPL labels. Extensive experiments on the variety of datasets (THUMOS '14, FineAction, GTEA, BEOID, and ActivityNet 1.3) demonstrate that the proposed approach is competitive with or outperforms prior methods for video-level and point-level supervision in terms of the trade-off between the annotation cost and detection performance.
format Preprint
id arxiv_https___arxiv_org_abs_2412_21205
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Action-Agnostic Point-Level Supervision for Temporal Action Detection
Yoshida, Shuhei M.
Shibata, Takashi
Terao, Makoto
Okatani, Takayuki
Sugiyama, Masashi
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
We propose action-agnostic point-level (AAPL) supervision for temporal action detection to achieve accurate action instance detection with a lightly annotated dataset. In the proposed scheme, a small portion of video frames is sampled in an unsupervised manner and presented to human annotators, who then label the frames with action categories. Unlike point-level supervision, which requires annotators to search for every action instance in an untrimmed video, frames to annotate are selected without human intervention in AAPL supervision. We also propose a detection model and learning method to effectively utilize the AAPL labels. Extensive experiments on the variety of datasets (THUMOS '14, FineAction, GTEA, BEOID, and ActivityNet 1.3) demonstrate that the proposed approach is competitive with or outperforms prior methods for video-level and point-level supervision in terms of the trade-off between the annotation cost and detection performance.
title Action-Agnostic Point-Level Supervision for Temporal Action Detection
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2412.21205