{"ID":2839051,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2511.16301","arxiv_id":"2511.16301","title":"Upsample Anything: A Simple and Hard to Beat Baseline for Feature Upsampling","abstract":"We present \\textbf{Upsample Anything}, a lightweight test-time optimization (TTO) framework that restores low-resolution features to high-resolution, pixel-wise outputs without any training. Although Vision Foundation Models demonstrate strong generalization across diverse downstream tasks, their representations are typically downsampled by 14x/16x (e.g., ViT), which limits their direct use in pixel-level applications. Existing feature upsampling approaches depend on dataset-specific retraining or heavy implicit optimization, restricting scalability and generalization. Upsample Anything addresses these issues through a simple per-image optimization that learns an anisotropic Gaussian kernel combining spatial and range cues, effectively bridging Gaussian Splatting and Joint Bilateral Upsampling. The learned kernel acts as a universal, edge-aware operator that transfers seamlessly across architectures and modalities, enabling precise high-resolution reconstruction of features, depth, or probability maps. It runs in only $\\approx0.419 \\text{s}$ per 224x224 image and achieves state-of-the-art performance on semantic segmentation, depth estimation, and both depth and probability map upsampling. \\textbf{Project page:} \\href{https://seominseok0429.github.io/Upsample-Anything/}{https://seominseok0429.github.io/Upsample-Anything/}","short_abstract":"We present \\textbf{Upsample Anything}, a lightweight test-time optimization (TTO) framework that restores low-resolution features to high-resolution, pixel-wise outputs without any training. Although Vision Foundation Models demonstrate strong generalization across diverse downstream tasks, their representations are ty...","url_abs":"https://arxiv.org/abs/2511.16301","url_pdf":"https://arxiv.org/pdf/2511.16301v2","authors":"[\"Minseok Seo\",\"Mark Hamilton\",\"Changick Kim\"]","published":"2025-11-20T12:27:53Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[]","has_code":false}