Saved in:
Bibliographic Details
Main Authors: Porto, Yannick, Martins, Renato, Chalumeau, Thomas, Demonceaux, Cedric
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.22695
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911705055887360
author Porto, Yannick
Martins, Renato
Chalumeau, Thomas
Demonceaux, Cedric
author_facet Porto, Yannick
Martins, Renato
Chalumeau, Thomas
Demonceaux, Cedric
contents Viewpoint change invariance and action temporal consistency are critical aspects for the effective deployment of human action detection of untrimmed videos. Existing appearance-based video detection methods often struggle with limited viewpoint diversity during training, while motion-based detection approaches frequently fail to model fine-grained temporal relationships across consecutive motion windows. This paper introduces a novel two-stage action detection approach designed to improve both view-invariance and global temporal coherence properties. In the first stage, we extract motion features from augmented virtual viewpoints, solely used at training. Then, the second stage introduces a new view-invariant, multi-scale temporal encoder based on selective state-space sequence modelling to aggregate information across viewpoints and time scales. Experiments on PKU-MMD and BABEL benchmarks demonstrate that this approach significantly outperforms state-of-the-art methods in all considered splits. Code and trained models are available at: https://icb-vision-ai.github.io/HydraView-TAD
format Preprint
id arxiv_https___arxiv_org_abs_2605_22695
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Improving Viewpoint-Invariance and Temporal Consistency for Action Detection
Porto, Yannick
Martins, Renato
Chalumeau, Thomas
Demonceaux, Cedric
Computer Vision and Pattern Recognition
Viewpoint change invariance and action temporal consistency are critical aspects for the effective deployment of human action detection of untrimmed videos. Existing appearance-based video detection methods often struggle with limited viewpoint diversity during training, while motion-based detection approaches frequently fail to model fine-grained temporal relationships across consecutive motion windows. This paper introduces a novel two-stage action detection approach designed to improve both view-invariance and global temporal coherence properties. In the first stage, we extract motion features from augmented virtual viewpoints, solely used at training. Then, the second stage introduces a new view-invariant, multi-scale temporal encoder based on selective state-space sequence modelling to aggregate information across viewpoints and time scales. Experiments on PKU-MMD and BABEL benchmarks demonstrate that this approach significantly outperforms state-of-the-art methods in all considered splits. Code and trained models are available at: https://icb-vision-ai.github.io/HydraView-TAD
title Improving Viewpoint-Invariance and Temporal Consistency for Action Detection
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2605.22695