---
title: "Towards Generalized Action Recognition on Low-Resolutions with Domain-Invariant Representation"
authors:
  - "Hao Li"
  - "Jinhui Xu"
  - "Dianlong You"
doi: "10.24963/ijcai.2026/146"
url: "https://ijcai.org/proceedings/2026/146"
pdf: "https://ijcai.org/proceedings/2026/0146.pdf"
venue: "Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence"
year: 2026
pages: "1306-1314"
track: "Main Track"
keywords:
  - "Computer Vision: Video analysis and understanding"
  - "Computer Vision: Action and behavior recognition"
  - "Computer Vision: Representation learning"
---

# Towards Generalized Action Recognition on Low-Resolutions with Domain-Invariant Representation

Hao Li, Jinhui Xu, Dianlong You

*Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence*, pages 1306-1314. DOI: [10.24963/ijcai.2026/146](https://doi.org/10.24963/ijcai.2026/146)

## Abstract

This paper studies cross-domain/species action recognition from low-resolution videos with sharp appearance variations and domain-specific biases. Its attractive viewpoint is mining domain-invariant representations of cross-domain/species features under rough spatial details to enhance recognition and generalization, overcoming the significant decline of existing methods. To address this, we propose a generalized Action recognition framework for Low-resolution conditions with Domain-invariant Representation learning, named ActLDR, designed to learn domain-invariant representations. First, it decomposes video understanding into spatial and temporal pathways for explicitly separating domain-dependent appearance cues from robust motion dynamics; Second, it constructs a Spatial-Temporal Feature Exchange module to enable cross-branch refinement and suppress domain bias; Third, we inject Gaussian feature interference to simulate feature corruption and enforce prediction-level consistency to encourage stable representations. Empirical results demonstrate that our proposal outperforms previous methods, significantly improving robustness across resolutions, domains, and species, and demonstrating outstanding generalization and transferability.