跳到正文
arXiv cs.AI· Hoang Pham Cong, Long D. Nguyen, Ho Viet Duc Luong, Nguyen Tran Minh Nhat, Binh P. Nguyen·· 2 小时前AI 评分43

DGSP:用假想物体探测规划器盲区的决策引导感知方法

Check Your Blind Spots: Decision-Guided Perception by Probing the Planner with Hypothetical Objects

AI 导读

论文提出决策引导假设感知方法 DGSP,把低分辨率场景当作草稿,向规划器插入假想物体探测哪些位置会影响决策,仅对这些位置和当前动作依赖的检测做高分辨率验证。作者证明在两类规划器上 DGSP 能精确复现全帧高分辨率感知的动作,并在三个驾驶数据集上用约四分之一的像素在约 98% 的帧上复现该动作,同时减少误刹车并找回更多决策相关物体。

正文 · 原文

View PDF HTML (experimental)

Abstract:Not every perception error changes the planner's action. Yet adaptive-resolution perception chooses where to spend high-resolution computation from image cues, which can neither reliably point to an object that low resolution missed nor tell which objects matter to the planner. We observe that the planner can say where an object would matter before anything is seen: inserting a hypothetical object at a location tests whether its action would change. Decision-Guided Speculative Perception (DGSP) treats the low-resolution scene and its action as a draft and probes the planner in this way to obtain a decision-sensitivity field, the locations where an unseen object would change the action. Only this field and the support, the detections on which the current action rests, are verified at high resolution, repeatedly until the action is stable. For two planner classes, we prove that DGSP returns exactly the action of whole-frame high-resolution perception if crop detections match whole-frame ones and the crops cover the field, conditions that real images meet approximately. On three driving datasets, it reproduces this action on about 98% of frames with about a quarter of the pixels. At the pixel cost of uniform downscaling, it repairs more of the draft's disagreements with this action, brakes falsely less often, and recovers many more of the decision-relevant objects that the draft misses.
Comments: Source Code: this https URL
Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
Cite as: arXiv:2611.23030 [cs.CV]
  (or arXiv:2611.23030v2 [cs.CV] for this version)
  https://doi.org/10.48550/arXiv.2611.23030

arXiv-issued DOI via DataCite

Submission history

From: Hoang Pham Cong [view email]
[v1] Tue, 17 Nov 2026 16:33:15 UTC (2,786 KB)
[v2] Wed, 18 Nov 2026 07:46:37 UTC (1,983 KB)

来源:arXiv cs.AI · arxiv.org