{"ID":2891945,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2507.18653","arxiv_id":"2507.18653","title":"Adapt, But Don't Forget: Fine-Tuning and Contrastive Routing for Lane Detection under Distribution Shift","abstract":"Lane detection models are often evaluated in a closed-world setting, where training and testing occur on the same dataset. We observe that, even within the same domain, cross-dataset distribution shifts can cause severe catastrophic forgetting during fine-tuning. To address this, we first train a base model on a source distribution and then adapt it to each new target distribution by creating separate branches, fine-tuning only selected components while keeping the original source branch fixed. Based on a component-wise analysis, we identify effective fine-tuning strategies for target distributions that enable parameter-efficient adaptation. At inference time, we propose using a supervised contrastive learning model to identify the input distribution and dynamically route it to the corresponding branch. Our framework achieves near-optimal F1-scores while using significantly fewer parameters than training separate models for each distribution.","short_abstract":"Lane detection models are often evaluated in a closed-world setting, where training and testing occur on the same dataset. We observe that, even within the same domain, cross-dataset distribution shifts can cause severe catastrophic forgetting during fine-tuning. To address this, we first train a base model on a source...","url_abs":"https://arxiv.org/abs/2507.18653","url_pdf":"https://arxiv.org/pdf/2507.18653v1","authors":"[\"Mohammed Abdul Hafeez Khan\",\"Parth Ganeriwala\",\"Sarah M. Lehman\",\"Siddhartha Bhattacharyya\",\"Amy Alvarez\",\"Natasha Neogi\"]","published":"2025-07-22T18:39:15Z","proceeding":"cs.CV","tasks":"[\"cs.CV\",\"cs.LG\"]","methods":"[]","has_code":false}