AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Scalable Place Recognition Under Appearance Change for Autonomous Driving

Anh-Dzung Doan, Yasir Latif, Tat-Jun Chin, Yu Liu, Thanh-Toan Do, Ian Reid

ICCV 2019 · Label-efficient and robust learning › Recognition and long-tail · approach: Correspondence and causal representation learning · 75 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 culling database update (ablation)HMM place recognition · Oxford RobotCar Seq-6mean translation error (m): 26.02m -> 18.88mImproved
add culling database update (ablation)HMM place recognition · Oxford RobotCar Seq-4mean translation error (m): 7.42m -> 5.80mImproved
add culling database update (ablation)HMM place recognition · Oxford RobotCar Seq-8median translation error (m): 13.28m -> 7.13mImproved
add combining of old places on top of culling (ablation)HMM place recognition · Oxford RobotCar Seq-4mean translation error (m): 5.80m -> 6.01mHurt
add combining of old places on top of culling (ablation)HMM place recognition · Oxford RobotCar Seq-7mean translation error (m): 30.06m -> 30.03mNo clear effect
add combining of old places on top of culling (ablation)HMM place recognition · Oxford RobotCar Seq-8inference time (ms/query): 3.91 -> 3.68Improved
add culling database update (ablation)HMM place recognition · Oxford RobotCar Seq-5inference time (ms/query): 4.24 -> 5.06Hurt
retrain MapNet on new query video with predicted labelsMapNet · Oxford RobotCar Seq-6mean translation error (m): 26.31m -> 28.69mHurt
retrain MapNet on new query video with predicted labelsMapNet · Oxford RobotCar Seq-4mean translation error (m): 8.92m -> 8.71mImproved
use HMM method with Cull+Combine instead of MapNetplace recognition · Oxford RobotCar Seq-6mean translation error (m): 26.31m -> 19.28mImproved
use HMM method instead of MapNetplace recognition · Oxford RobotCar Seq-4median translation error (m): 4.53m -> 5.65mHurt
use HMM method instead of MapNet trainingplace recognition · Oxford RobotCar Seq-1,2training time: 11.6h -> 98.9sImproved

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