{"ID":2838013,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2511.18452","arxiv_id":"2511.18452","title":"NAF: Zero-Shot Feature Upsampling via Neighborhood Attention Filtering","abstract":"Vision Foundation Models (VFMs) extract spatially downsampled representations, posing challenges for pixel-level tasks. Existing upsampling approaches face a fundamental trade-off: classical filters are fast and broadly applicable but rely on fixed forms, while modern upsamplers achieve superior accuracy through learnable, VFM-specific forms at the cost of retraining for each VFM. We introduce Neighborhood Attention Filtering (NAF), which bridges this gap by learning adaptive spatial-and-content weights through Cross-Scale Neighborhood Attention and Rotary Position Embeddings (RoPE), guided solely by the high-resolution input image. NAF operates zero-shot: it upsamples features from any VFM without retraining, making it the first VFM-agnostic architecture to outperform VFM-specific upsamplers and achieve state-of-the-art performance across multiple downstream tasks. It maintains high efficiency, scaling to 2K feature maps and reconstructing intermediate-resolution maps at 18 FPS. Beyond feature upsampling, NAF demonstrates strong performance on image restoration, highlighting its versatility. Code and checkpoints are available at https://github.com/valeoai/NAF.","short_abstract":"Vision Foundation Models (VFMs) extract spatially downsampled representations, posing challenges for pixel-level tasks. Existing upsampling approaches face a fundamental trade-off: classical filters are fast and broadly applicable but rely on fixed forms, while modern upsamplers achieve superior accuracy through learna...","url_abs":"https://arxiv.org/abs/2511.18452","url_pdf":"https://arxiv.org/pdf/2511.18452v1","authors":"[\"Loick Chambon\",\"Paul Couairon\",\"Eloi Zablocki\",\"Alexandre Boulch\",\"Nicolas Thome\",\"Matthieu Cord\"]","published":"2025-11-23T13:43:52Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[]","has_code":false,"code_links":[{"ID":606730,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2838013,"paper_url":"https://arxiv.org/abs/2511.18452","paper_title":"NAF: Zero-Shot Feature Upsampling via Neighborhood Attention Filtering","repo_url":"https://github.com/valeoai/NAF","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}