Towards Fine-Grained Emotion Understanding via Skeleton-Based Micro-Gesture Recognition

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Xu, Hao, Cheng, Lechao, Wang, Yaxiong, Tang, Shengeng, Zhong, Zhun
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915345957126144
author Xu, Hao
Cheng, Lechao
Wang, Yaxiong
Tang, Shengeng
Zhong, Zhun
author_facet Xu, Hao
Cheng, Lechao
Wang, Yaxiong
Tang, Shengeng
Zhong, Zhun
contents We present our solution to the MiGA Challenge at IJCAI 2025, which aims to recognize micro-gestures (MGs) from skeleton sequences for the purpose of hidden emotion understanding. MGs are characterized by their subtlety, short duration, and low motion amplitude, making them particularly challenging to model and classify. We adopt PoseC3D as the baseline framework and introduce three key enhancements: (1) a topology-aware skeleton representation specifically designed for the iMiGUE dataset to better capture fine-grained motion patterns; (2) an improved temporal processing strategy that facilitates smoother and more temporally consistent motion modeling; and (3) the incorporation of semantic label embeddings as auxiliary supervision to improve the model generalization. Our method achieves a Top-1 accuracy of 67.01\% on the iMiGUE test set. As a result of these contributions, our approach ranks third on the official MiGA Challenge leaderboard. The source code is available at \href{https://github.com/EGO-False-Sleep/Miga25_track1}{https://github.com/EGO-False-Sleep/Miga25\_track1}.
format Preprint
id arxiv_https___arxiv_org_abs_2506_12848
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Towards Fine-Grained Emotion Understanding via Skeleton-Based Micro-Gesture Recognition
Xu, Hao
Cheng, Lechao
Wang, Yaxiong
Tang, Shengeng
Zhong, Zhun
Computer Vision and Pattern Recognition
We present our solution to the MiGA Challenge at IJCAI 2025, which aims to recognize micro-gestures (MGs) from skeleton sequences for the purpose of hidden emotion understanding. MGs are characterized by their subtlety, short duration, and low motion amplitude, making them particularly challenging to model and classify. We adopt PoseC3D as the baseline framework and introduce three key enhancements: (1) a topology-aware skeleton representation specifically designed for the iMiGUE dataset to better capture fine-grained motion patterns; (2) an improved temporal processing strategy that facilitates smoother and more temporally consistent motion modeling; and (3) the incorporation of semantic label embeddings as auxiliary supervision to improve the model generalization. Our method achieves a Top-1 accuracy of 67.01\% on the iMiGUE test set. As a result of these contributions, our approach ranks third on the official MiGA Challenge leaderboard. The source code is available at \href{https://github.com/EGO-False-Sleep/Miga25_track1}{https://github.com/EGO-False-Sleep/Miga25\_track1}.
title Towards Fine-Grained Emotion Understanding via Skeleton-Based Micro-Gesture Recognition
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2506.12848