<strong>Learning A Unified Risk Map for Autonomous Driving in Partially Observable Environments (IEEE RA-L)</strong> — How should a self-driving car reason about danger it can't see? This work (IEEE RA-L) introduces a unified spatiotemporal risk map that fuses traffic-flow risk and collision risk into a single field for occlusion-aware planning. Its standout idea: generate rich risk labels offline by combining an expressive guided-diffusion model with rule-based logic, then distill them into a lightweight transformer that predicts the field in real time from vectorized maps and the ego's field-of-view — the richness of a generative model at feed-forward speed. To beat the scarcity of safety-critical occluded data, a guidance-based diffusion generator synthesizes adversarial-yet-realistic interactions. On the Waymo Open Motion Dataset, it delivers consistent safety gains (e.g., TTC) over SOTA occlusion-aware baselines — a practical \"use a heavy model to teach a fast one\" template for safety-critical robotics.</p>\n","updatedAt":"2026-05-29T03:18:22.218Z","author":{"_id":"646eac510867c99c2d3fde08","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/646eac510867c99c2d3fde08/fvIiW7zj4aTNbp16kTNBA.jpeg","fullname":"Yaofeng Su","name":"Exploration","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.8654411435127258},"editors":["Exploration"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/646eac510867c99c2d3fde08/fvIiW7zj4aTNbp16kTNBA.jpeg"],"reactions":[],"isReport":false}},{"id":"6a1a40ba89a9cfd05a97eb3e","author":{"_id":"63d3e0e8ff1384ce6c5dd17d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg","fullname":"Librarian Bot (Bot)","name":"librarian-bot","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":359,"isUserFollowing":false},"createdAt":"2026-05-30T01:43:22.000Z","type":"comment","data":{"edited":false,"hidden":false,"latest":{"raw":"This is an automated message from the [Librarian Bot](https://huggingface.co/librarian-bots). I found the following papers similar to this paper. \n\nThe following papers were recommended by the Semantic Scholar API \n\n* [SparseWorld: Enhancing End-to-End Autonomous Driving via World Models with Sparse Scene Representation](https://huggingface.co/papers/2605.24354) (2026)\n* [ExploreVLA: Dense World Modeling and Exploration for End-to-End Autonomous Driving](https://huggingface.co/papers/2604.02714) (2026)\n* [SAIL: Scene-aware adaptive iterative learning for long-tail trajectory prediction in autonomous vehicles](https://huggingface.co/papers/2604.04573) (2026)\n* [Not All Agents Matter: From Global Attention Dilution to Risk-Prioritized Game Planning](https://huggingface.co/papers/2604.05449) (2026)\n* [MAPLE: Latent Multi-Agent Play for End-to-End Autonomous Driving](https://huggingface.co/papers/2605.14201) (2026)\n* [Beyond Imitation: Learning Safe End-to-End Autonomous Driving from Hard Negatives](https://huggingface.co/papers/2605.19771) (2026)\n* [MISTY: High-Throughput Motion Planning via Mixer-based Single-step Drifting](https://huggingface.co/papers/2604.21489) (2026)\n\n\n Please give a thumbs up to this comment if you found it helpful!\n\n If you want recommendations for any Paper on Hugging Face checkout [this](https://huggingface.co/spaces/librarian-bots/recommend_similar_papers) Space\n\n You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: `@librarian-bot recommend`","html":"<p>This is an automated message from the <a href=\"https://huggingface.co/librarian-bots\">Librarian Bot</a>. I found the following papers similar to this paper. </p>\n<p>The following papers were recommended by the Semantic Scholar API </p>\n<ul>\n<li><a href=\"https://huggingface.co/papers/2605.24354\">SparseWorld: Enhancing End-to-End Autonomous Driving via World Models with Sparse Scene Representation</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2604.02714\">ExploreVLA: Dense World Modeling and Exploration for End-to-End Autonomous Driving</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2604.04573\">SAIL: Scene-aware adaptive iterative learning for long-tail trajectory prediction in autonomous vehicles</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2604.05449\">Not All Agents Matter: From Global Attention Dilution to Risk-Prioritized Game Planning</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2605.14201\">MAPLE: Latent Multi-Agent Play for End-to-End Autonomous Driving</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2605.19771\">Beyond Imitation: Learning Safe End-to-End Autonomous Driving from Hard Negatives</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2604.21489\">MISTY: High-Throughput Motion Planning via Mixer-based Single-step Drifting</a> (2026)</li>\n</ul>\n<p> Please give a thumbs up to this comment if you found it helpful!</p>\n<p> If you want recommendations for any Paper on Hugging Face checkout <a href=\"https://huggingface.co/spaces/librarian-bots/recommend_similar_papers\">this</a> Space</p>\n<p> You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: <code><span class=\"SVELTE_PARTIAL_HYDRATER contents\" data-target=\"UserMention\" data-props=\"{"user":"librarian-bot"}\"><span class=\"inline-block\"><span class=\"contents\"><a href=\"/librarian-bot\">@<span class=\"underline\">librarian-bot</span></a></span> </span></span> recommend</code></p>\n","updatedAt":"2026-05-30T01:43:22.090Z","author":{"_id":"63d3e0e8ff1384ce6c5dd17d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg","fullname":"Librarian Bot (Bot)","name":"librarian-bot","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":359,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.7381721138954163},"editors":["librarian-bot"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2605.22189","authors":[{"_id":"6a1835f86b5dee569682b557","name":"Jie Jia","hidden":false},{"_id":"6a1835f86b5dee569682b558","name":"Yaofeng Su","hidden":false},{"_id":"6a1835f86b5dee569682b559","name":"Zeyu Bao","hidden":false},{"_id":"6a1835f86b5dee569682b55a","name":"Yun Hong","hidden":false},{"_id":"6a1835f86b5dee569682b55b","name":"Bingzhao Gao","hidden":false},{"_id":"6a1835f86b5dee569682b55c","name":"Zhongxue Gan","hidden":false},{"_id":"6a1835f86b5dee569682b55d","name":"Wenchao Ding","hidden":false}],"publishedAt":"2026-05-21T00:00:00.000Z","submittedOnDailyAt":"2026-05-29T00:00:00.000Z","title":"Learning A Unified Risk Map for Autonomous Driving in Partially Observable Environments","submittedOnDailyBy":{"_id":"646eac510867c99c2d3fde08","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/646eac510867c99c2d3fde08/fvIiW7zj4aTNbp16kTNBA.jpeg","isPro":false,"fullname":"Yaofeng Su","user":"Exploration","type":"user","name":"Exploration"},"summary":"Occlusion-aware prediction remains a critical challenge in autonomous driving due to the inherent uncertainty of unobserved regions. Existing approaches either overestimate risk based on reachable states or struggle to predict accurate trajectories under high occlusion uncertainty. To address these limitations, we propose a unified risk map modeling and learning framework for partially observable environments. Our method integrates traffic flow risk and collision risk through spatiotemporal modeling, enabling fine-grained assessment of occlusion-induced hazards. To address the scarcity of scenarios involving occluded interactions, we introduce a diffusion-based scenario generation framework that produces realistic yet adversarial scenarios. We integrate the modeling and learning of a unified risk map into a framework that supports risk-aware planning under partial observability. Experiments on the Waymo Open Motion Dataset show that our method significantly outperforms the state-of-the-art occlusion-aware baseline, improving minimum time-to-collision by 0.78 times and average time-to-collision by 1.67 times. The proposed framework offers a comprehensive and practical solution for risk-aware planning in partially observable environments.","upvotes":4,"discussionId":"6a1835f96b5dee569682b55e","ai_summary":"A unified risk map modeling framework addresses occlusion challenges in autonomous driving by integrating traffic flow and collision risks through spatiotemporal modeling and diffusion-based scenario generation.","ai_keywords":["risk map modeling","spatiotemporal modeling","occlusion-aware prediction","diffusion-based scenario generation","risk-aware planning","partially observable environments"],"organization":{"_id":"643cb0625fcffe09fb6ca688","name":"Fudan-University","fullname":"Fudan University","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/6437eca0819f3ab20d162e14/kWv0cGlAhAG3iNWVxowkJ.png"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"646eac510867c99c2d3fde08","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/646eac510867c99c2d3fde08/fvIiW7zj4aTNbp16kTNBA.jpeg","isPro":false,"fullname":"Yaofeng Su","user":"Exploration","type":"user"},{"_id":"695dc4e545f7a34d8e089e03","avatarUrl":"/avatars/e4e772140374aabf4e555b77ae95e484.svg","isPro":false,"fullname":"Shuwei Liu","user":"LSW628","type":"user"},{"_id":"68acd0128711ce479b20f7df","avatarUrl":"/avatars/10e7f2ddfecd955be69e500af2169514.svg","isPro":false,"fullname":"Yun Hong","user":"yunH16","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"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"643cb0625fcffe09fb6ca688","name":"Fudan-University","fullname":"Fudan University","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/6437eca0819f3ab20d162e14/kWv0cGlAhAG3iNWVxowkJ.png"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2605/2605.22189.md"}">
Learning A Unified Risk Map for Autonomous Driving in Partially Observable Environments
Abstract
A unified risk map modeling framework addresses occlusion challenges in autonomous driving by integrating traffic flow and collision risks through spatiotemporal modeling and diffusion-based scenario generation.
AI-generated summary
Occlusion-aware prediction remains a critical challenge in autonomous driving due to the inherent uncertainty of unobserved regions. Existing approaches either overestimate risk based on reachable states or struggle to predict accurate trajectories under high occlusion uncertainty. To address these limitations, we propose a unified risk map modeling and learning framework for partially observable environments. Our method integrates traffic flow risk and collision risk through spatiotemporal modeling, enabling fine-grained assessment of occlusion-induced hazards. To address the scarcity of scenarios involving occluded interactions, we introduce a diffusion-based scenario generation framework that produces realistic yet adversarial scenarios. We integrate the modeling and learning of a unified risk map into a framework that supports risk-aware planning under partial observability. Experiments on the Waymo Open Motion Dataset show that our method significantly outperforms the state-of-the-art occlusion-aware baseline, improving minimum time-to-collision by 0.78 times and average time-to-collision by 1.67 times. The proposed framework offers a comprehensive and practical solution for risk-aware planning in partially observable environments.
Community
Learning A Unified Risk Map for Autonomous Driving in Partially Observable Environments (IEEE RA-L) — How should a self-driving car reason about danger it can't see? This work (IEEE RA-L) introduces a unified spatiotemporal risk map that fuses traffic-flow risk and collision risk into a single field for occlusion-aware planning. Its standout idea: generate rich risk labels offline by combining an expressive guided-diffusion model with rule-based logic, then distill them into a lightweight transformer that predicts the field in real time from vectorized maps and the ego's field-of-view — the richness of a generative model at feed-forward speed. To beat the scarcity of safety-critical occluded data, a guidance-based diffusion generator synthesizes adversarial-yet-realistic interactions. On the Waymo Open Motion Dataset, it delivers consistent safety gains (e.g., TTC) over SOTA occlusion-aware baselines — a practical "use a heavy model to teach a fast one" template for safety-critical robotics.
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on Hugging Face checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: @librarian-bot recommend
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/2605.22189 in a model README.md to link it from this page.
Cite arxiv.org/abs/2605.22189 in a dataset README.md to link it from this page.
Cite arxiv.org/abs/2605.22189 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.