HEDP: A Hybrid Energy-Distance Prompt-based Framework for Domain Incremental Learning
Abstract
Domain Incremental Learning is a critical scenario that requires models to continuously adapt to new data domains without retraining. However, domain shifts often cause severe performance degradation. To address this, we propose Hybrid Energy-Distance Prompt, a domain-incremental framework inspired by Helmholtz free energy. HEDP introduces an energy regularization loss to enhance the separability of domain representations and a hybrid energy-distance weighted mechanism that fuses energy-based and distance-based cues to improve domain selection and generalization. Experiments on multiple benchmarks, including CORe50, show that HEDP achieves superior performance on unseen domains with a 2.57\% accuracy gain, effectively mitigating catastrophic forgetting and enhancing open-world adaptability. Our code is \href{https://github.com/dannis97500/HEDP/}{available here}.
Cite
@article{arxiv.2605.05776,
title = {HEDP: A Hybrid Energy-Distance Prompt-based Framework for Domain Incremental Learning},
author = {Yu Feng and Zhen Tian and Haoran Luo and Xie Yu and Diancheng Cheng and Haoyue Zheng and Shuai Lyu and Ping Zong and Lianyuan Li and Xin Ge and Yifan Zhu},
journal= {arXiv preprint arXiv:2605.05776},
year = {2026}
}
Comments
13 pages, 6 figures, Accepted by ICML 2026