English

Minimizing Embedding Distortion for Robust Out-of-Distribution Performance

Computer Vision and Pattern Recognition 2024-09-13 v1

Abstract

Foundational models, trained on vast and diverse datasets, have demonstrated remarkable capabilities in generalizing across different domains and distributions for various zero-shot tasks. Our work addresses the challenge of retaining these powerful generalization capabilities when adapting foundational models to specific downstream tasks through fine-tuning. To this end, we introduce a novel approach we call "similarity loss", which can be incorporated into the fine-tuning process of any task. By minimizing the distortion of fine-tuned embeddings from the pre-trained embeddings, our method strikes a balance between task-specific adaptation and preserving broad generalization abilities. We evaluate our approach on two diverse tasks: image classification on satellite imagery and face recognition, focusing on open-class and domain shift scenarios to assess out-of-distribution (OOD) performance. We demonstrate that this approach significantly improves OOD performance while maintaining strong in-distribution (ID) performance.

Keywords

Cite

@article{arxiv.2409.07582,
  title  = {Minimizing Embedding Distortion for Robust Out-of-Distribution Performance},
  author = {Tom Shaked and Yuval Goldman and Oran Shayer},
  journal= {arXiv preprint arXiv:2409.07582},
  year   = {2024}
}

Comments

Accepted to ECCV 2024 workshop

R2 v1 2026-06-28T18:41:45.950Z