AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

Testing the Safety of Self-driving Vehicles by Simulating Perception and Prediction

Kelvin Wong, Qiang Zhang, Ming Liang, Bin Yang, Renjie Liao, Abbas Sadat, Raquel Urtasun

ECCV 2020 · Probabilistic and scientific modeling › Molecules and proteins · approach: Multimodal trajectory prediction · 21 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 learned perception/prediction noise (ContextNoise) to perfect-perception simulationmotion planning simulation · structured test track vehicleAP @0.5 IoU: 77.2 -> 94.5Improved
add ContextNoisesimulation · structured test track pedestrianAP @0.3 IoU: 47.4 -> 81.9Improved
add ContextNoisesimulation · structured test track bicyclistAP @0.3 IoU: 47.8 -> 89.6Improved
add ContextNoisePLT planner simulation · structured test trackcollision simulation IoU: 70.3 -> 85.2Improved
add ContextNoiseACC planner simulation · structured test trackl2 distance at 3.0s: 19.8 -> 15.1Improved
add ContextNoisesimulation · structured test track pedestrianADE at 60% recall: 36 -> 37No clear effect
add ContextNoisesimulation · structured test track bicyclistADE at 50% recall: 77 -> 79Hurt
increase ContextNoise training data from 5% to 100% of TOR4D train split (ablation)ContextNoise · TOR4D val vehicleFDE at 70% recall: 123 -> 70Improved
increase ContextNoise training data from 5% to 100% (ablation)ContextNoise · TOR4D val vehicleAP @0.5 IoU: 88.0 -> 91.7Improved
halve ContextNoise training data (100% to 50%) (ablation)ContextNoise PLT · TOR4D valcollision simulation IoU: 74.9 -> 74.0Hurt
halve ContextNoise training data (100% to 50%) (ablation)ContextNoise ACC · TOR4D valcollision simulation IoU: 64.3 -> 63.1Hurt
train ContextNoise on only 5% of data (ablation)ContextNoise · TOR4D val bicyclistADE at 50% recall: 49 -> 126Hurt

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