AI Paper Atlas

Research map / Detection and segmentation / 3D detection and driving

Test-Time Correction: An Online 3D Detection System via Visual Prompting

Hanxue Zhang, Zetong Yang, Yanan Sun, Li‐Qun Chen, Fei Xia, Fatma Güney, Hongyang Li

TPAMI 2025 · Detection and segmentation › 3D detection and driving · approach: LiDAR perception for driving · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add Test-time Correction with visual prompts at 5-frame feedbackBEVFormer · nuScenes valmAP (%): 41.1 -> 50.2Improved
add Test-time Correction with visual promptsStreamPETR · nuScenes valmAP (%): 45.0 -> 51.6Improved
add Test-time Correction to the monocular detectorMonoDETR · nuScenes valmAP (%): 42.6 -> 45.0Improved
add TTC to a detector trained with 20% distant labelsMonoDETR · nuScenes distant objectsmAP (%): 0.0 -> 10.1Improved
add TTC to a detector trained with 20% vehicle labelsMonoDETR · nuScenes vehiclesmAP (%): 2.4 -> 22.4Improved
add TTC for unseen classes (zero-shot)MonoDETR · nuScenes unseen classesmAP (%): 0.0 -> 7.2Improved
add TTC under rain and night domain shiftMonoDETR · nuScenes rain and nightmAP (%): 25.2 -> 29.0Improved
add explicit similarity loss (Focal and Dice) to alignment (ablation)TTC-MonoDETR · nuScenesmAP (%): 32.6 -> 39.4Improved
add localization loss on top of similarity loss (ablation)TTC-MonoDETR · nuScenesmAP (%): 39.4 -> 43.3Improved
predict N=4 positions per visual prompt instead of 1 (ablation)TTC-MonoDETR · nuScenesmAP (%): 39.9 -> 43.3Improved
increase N from 4 to 8 positions (ablation)TTC-MonoDETR · nuScenesmAP (%): 43.3 -> 43.0No clear effect
replace OA visual prompts with SOT tracking (ATOM) plus 3D headMV2D · nuScenes valmAP (%): 42.6 -> 42.1Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy