DARA: Degradation-Aware Low-Rank Residual Adaptation with Original-to-Corrupted Distillation for Corruption-Robust Animal Re-Identification
Computer Vision and Pattern Recognition
Summary
The authors looked at how animal re-identification (Re-ID), which helps recognize individual animals, can get worse when images are blurry or noisy. They created a method called DARA that fixes the features used to identify animals when the images are degraded, without needing labels about the type of corruption. DARA keeps the main model unchanged and adds a small module to adjust the features, learning from clear images to keep the identity correct. Tests showed DARA helps the system work better on poor-quality images and new types of distortions with very little extra computing cost.
Authors
Cynthia Xie, Talia Xu
Abstract
Animal re-identification (Re-ID) relies on fine-grained identity cues that can be disrupted by blur, noise, compression, and other visual degradations. Existing robustness strategies based on degradation-augmented training or pixel-level restoration improve robustness indirectly, but do not explicitly repair shifts in the identity retrieval space. We study corruption-robust animal Re-ID as input-conditioned feature-space repair and introduce DARA, a lightweight retrofit for compact Re-ID models. DARA freezes the fine-tuned backbone and learns routed low-rank residual experts to adapt degraded-input embeddings without corruption-type annotations. To stabilize this adaptive repair, original-to-corrupted distillation uses an original-image teacher to preserve individual embeddings and retrieval relations. Experiments on ATRW, FriesianCattle2017, MPDD, and SeaStarReID2023 show that DARA improves corrupted-query retrieval over standard and augmentation-based fine-tuning, generalizes to unseen corruptions and cross-domain evaluation, and recovers 77.0% of the corrupted-query mAP gap to full corrupted fine-tuning while adding only 0.49% parameters and 0.05% FLOPs.