Submission details of EPNet_lidar

Name EPNet_lidar
Paper Link
Code Link
AP 0.592524
Input N/A
Runtime 0.23 s
Environment 1 GPU(Titan V)
Abstract In this paper, we aim at addressing two critical issues in the 3D detection task, including the exploitation of multiple sensors (namely LiDAR point cloud and camera image), as well as the inconsistency between the localization and classification confidence.To this end, we propose a novel fusion module to enhance the point features with semantic image features in a point-wise manner without any image annotations. Besides, a consistency enforcing loss is employed to explicitly encourage the consistency of both the localization and classification confidence.

Detailed results

Overall precision/recall curve

Per-sequence results

sequence name AP
cubberly-auditorium-2019-04-22_1 0.755045
discovery-walk-2019-02-28_0 0.887992
discovery-walk-2019-02-28_1 0.849797
food-trucks-2019-02-12_0 0.757168
gates-ai-lab-2019-04-17_0 0.687834
gates-basement-elevators-2019-01-17_0 0.881297
gates-foyer-2019-01-17_0 0.841168
gates-to-clark-2019-02-28_0 0.89543
hewlett-class-2019-01-23_0 0.738318
hewlett-class-2019-01-23_1 0.944201
huang-2-2019-01-25_1 0.682878
huang-intersection-2019-01-22_0 0.832757
indoor-coupa-cafe-2019-02-06_0 0.554418
lomita-serra-intersection-2019-01-30_0 0.938251
meyer-green-2019-03-16_1 0.647129
nvidia-aud-2019-01-25_0 0.616565
nvidia-aud-2019-04-18_1 0.681321
nvidia-aud-2019-04-18_2 0.770224
outdoor-coupa-cafe-2019-02-06_0 0.636101
quarry-road-2019-02-28_0 0.781164
serra-street-2019-01-30_0 0.887442
stlc-111-2019-04-19_1 0.901797
stlc-111-2019-04-19_2 0.811985
tressider-2019-03-16_2 0.810315
tressider-2019-04-26_0 0.466559
tressider-2019-04-26_1 0.471544
tressider-2019-04-26_3 0.502213
total 0.592524