2</sup>FIX (Distractor-Free DIFIX), a diffusion-based enhancement module that improves radiance field renderings.","html":"<p>DF3DV-1K, a large-scale real-world dataset for distractor-free novel view synthesis, comprising 1,000+ scenes with clean and cluttered images per scene, together with DI<sup>2</sup>FIX (Distractor-Free DIFIX), a diffusion-based enhancement module that improves radiance field renderings.</p>\n","updatedAt":"2026-06-19T06:21:10.692Z","author":{"_id":"670753680681f4d0a94ebccf","avatarUrl":"/avatars/1aa6f063bacdb25d36784d0f93bb2224.svg","fullname":"ChengYou Lu","name":"ChengYou305","type":"user","isPro":true,"isHf":false,"isHfAdmin":false,"isMod":false,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.816436767578125},"editors":["ChengYou305"],"editorAvatarUrls":["/avatars/1aa6f063bacdb25d36784d0f93bb2224.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2604.13416","authors":[{"_id":"6a34dde34c5c5e0d69bf1d80","name":"Cheng-You Lu","hidden":false},{"_id":"6a34dde34c5c5e0d69bf1d81","name":"Yi-Shan Hung","hidden":false},{"_id":"6a34dde34c5c5e0d69bf1d82","name":"Wei-Ling Chi","hidden":false},{"_id":"6a34dde34c5c5e0d69bf1d83","name":"Hao-Ping Wang","hidden":false},{"_id":"6a34dde34c5c5e0d69bf1d84","name":"Charlie Li-Ting Tsai","hidden":false},{"_id":"6a34dde34c5c5e0d69bf1d85","name":"Yu-Cheng Chang","hidden":false},{"_id":"6a34dde34c5c5e0d69bf1d86","name":"Yu-Lun Liu","hidden":false},{"_id":"6a34dde34c5c5e0d69bf1d87","name":"Thomas Do","hidden":false},{"_id":"6a34dde34c5c5e0d69bf1d88","name":"Chin-Teng Lin","hidden":false}],"mediaUrls":["https://cdn-uploads.huggingface.co/production/uploads/670753680681f4d0a94ebccf/rX-HcnsdAk075HsAnFZG-.mp4"],"publishedAt":"2026-06-18T00:00:00.000Z","submittedOnDailyAt":"2026-06-19T00:00:00.000Z","title":"DF3DV-1K: A Large-Scale Dataset and Benchmark for Distractor-Free Novel View Synthesis","submittedOnDailyBy":{"_id":"670753680681f4d0a94ebccf","avatarUrl":"/avatars/1aa6f063bacdb25d36784d0f93bb2224.svg","isPro":true,"fullname":"ChengYou Lu","user":"ChengYou305","type":"user","name":"ChengYou305"},"summary":"Advances in radiance fields have enabled photorealistic novel view synthesis. In several domains, large-scale real-world datasets have been developed to support comprehensive benchmarking and to facilitate progress beyond scene-specific reconstruction. However, for distractor-free radiance fields, a large-scale dataset with clean and cluttered images per scene remains lacking, limiting the development. To address this gap, we introduce DF3DV-1K, a large-scale real-world dataset comprising 1,048 scenes, each providing clean and cluttered image sets for benchmarking. In total, the dataset contains 89,924 images captured using consumer cameras to mimic casual capture, spanning 128 distractor types and 161 scene themes across indoor and outdoor environments. A curated subset of 41 scenes, DF3DV-41, is systematically designed to evaluate the robustness of distractor-free radiance field methods under challenging scenarios. Using DF3DV-1K, we benchmark nine recent distractor-free radiance field methods and 3D Gaussian Splatting, identifying the most robust methods and the most challenging scenarios. Beyond benchmarking, we demonstrate an application of DF3DV-1K by fine-tuning a diffusion-based 2D enhancer to improve radiance field methods, achieving average improvements of 0.96 dB PSNR and 0.057 LPIPS on the held-out set (e.g., DF3DV-41) and the On-the-go dataset. We hope DF3DV-1K facilitates the development of distractor-free vision and promotes progress beyond scene-specific approaches. The dataset and leaderboard are available at https://johnnylu305.github.io/df3dv1k_web/.","upvotes":4,"discussionId":"6a34dde44c5c5e0d69bf1d89","projectPage":"https://johnnylu305.github.io/df3dv1k_web/","githubRepo":"https://github.com/johnnylu305/DF3DV","githubRepoAddedBy":"user","ai_summary":"A large-scale real-world dataset called DF3DV-1K is introduced to address the lack of clean and cluttered image sets for distractor-free radiance field research, containing 1,048 scenes with 89,924 images across 128 distractor types and 161 scene themes, along with a curated subset DF3DV-41 for robustness evaluation, and demonstrates improved performance when used to fine-tune a diffusion-based 2D enhancer for radiance field methods.","ai_keywords":["radiance fields","distractor-free","novel view synthesis","diffusion-based 2D enhancer","3D Gaussian Splatting","PSNR","LPIPS"],"ai_summary_model":"Qwen/Qwen2.5-Coder-32B-Instruct","githubStars":9},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"670753680681f4d0a94ebccf","avatarUrl":"/avatars/1aa6f063bacdb25d36784d0f93bb2224.svg","isPro":true,"fullname":"ChengYou Lu","user":"ChengYou305","type":"user"},{"_id":"619f9755da83161f25840698","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/619f9755da83161f25840698/FM421pE1mz5v1YhrxA8ZA.jpeg","isPro":false,"fullname":"Muhammad Umair","user":"umair894","type":"user"},{"_id":"68246c555c53974b2d293d92","avatarUrl":"/avatars/fd375f5bcc7835eed58e1e0367f05285.svg","isPro":false,"fullname":"Xiaowei Jiang","user":"xynico","type":"user"},{"_id":"6a34ed8f7a78da7522a49916","avatarUrl":"/avatars/47c74624d342eb3e0a11963aaa31f3c9.svg","isPro":false,"fullname":"Jie Yang","user":"Boblazar88","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"query":{}}">
DF3DV-1K: A Large-Scale Dataset and Benchmark for Distractor-Free Novel View Synthesis
Abstract
A large-scale real-world dataset called DF3DV-1K is introduced to address the lack of clean and cluttered image sets for distractor-free radiance field research, containing 1,048 scenes with 89,924 images across 128 distractor types and 161 scene themes, along with a curated subset DF3DV-41 for robustness evaluation, and demonstrates improved performance when used to fine-tune a diffusion-based 2D enhancer for radiance field methods.
Advances in radiance fields have enabled photorealistic novel view synthesis. In several domains, large-scale real-world datasets have been developed to support comprehensive benchmarking and to facilitate progress beyond scene-specific reconstruction. However, for distractor-free radiance fields, a large-scale dataset with clean and cluttered images per scene remains lacking, limiting the development. To address this gap, we introduce DF3DV-1K, a large-scale real-world dataset comprising 1,048 scenes, each providing clean and cluttered image sets for benchmarking. In total, the dataset contains 89,924 images captured using consumer cameras to mimic casual capture, spanning 128 distractor types and 161 scene themes across indoor and outdoor environments. A curated subset of 41 scenes, DF3DV-41, is systematically designed to evaluate the robustness of distractor-free radiance field methods under challenging scenarios. Using DF3DV-1K, we benchmark nine recent distractor-free radiance field methods and 3D Gaussian Splatting, identifying the most robust methods and the most challenging scenarios. Beyond benchmarking, we demonstrate an application of DF3DV-1K by fine-tuning a diffusion-based 2D enhancer to improve radiance field methods, achieving average improvements of 0.96 dB PSNR and 0.057 LPIPS on the held-out set (e.g., DF3DV-41) and the On-the-go dataset. We hope DF3DV-1K facilitates the development of distractor-free vision and promotes progress beyond scene-specific approaches. The dataset and leaderboard are available at https://johnnylu305.github.io/df3dv1k_web/.
Community
DF3DV-1K, a large-scale real-world dataset for distractor-free novel view synthesis, comprising 1,000+ scenes with clean and cluttered images per scene, together with DI2FIX (Distractor-Free DIFIX), a diffusion-based enhancement module that improves radiance field renderings.
Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images
Cite arxiv.org/abs/2604.13416 in a model README.md to link it from this page.
Cite arxiv.org/abs/2604.13416 in a dataset README.md to link it from this page.
Cite arxiv.org/abs/2604.13416 in a Space README.md to link it from this page.
Discussion (0)
Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.
Sign in →No comments yet. Sign in and be the first to say something.