AI Paper Atlas

Research map / Point clouds / Segmentation and detection

LidarNAS: Unifying and Searching Neural Architectures for 3D Point Clouds

Chenxi Liu, Zhaoqi Leng, Pei Sun, Shuyang Cheng, Charles R. Qi, Yin Zhou, Mingxing Tan, Dragomir Anguelov

ECCV 2022 · Point clouds › Segmentation and detection · approach: LiDAR voxel and pillar detectors

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
evolve architecture with LidarNAS search from PointPillars-likePointPillars-like · Waymo vehicleL1 3D AP: 67.6 -> 73.2Improved
evolve architecture with LidarNAS search from RSNRSN · Waymo vehicleL1 3D AP: 75.2 -> 75.6Improved
evolve architecture with LidarNAS search from RSNRSN · Waymo pedestrianL1 3D AP: 77.1 -> 77.4Improved
add 2D U-Net on range image before voxelization to pillar view (ablation)PointPillars-like · Waymo vehicleL1 3D AP: 67.6 -> 72.3Improved
decrease voxel size from 0.32 to 0.25 (ablation)LidarNASNet-P · Waymo vehicleL1 3D AP: 72.3 -> 73.2Improved
widen RSN sparse U-Net channels from 64 to 91 instead of searching (ablation)RSN · Waymo vehiclelatency (ms): 60.8 ms vs 49.3 msHurt
add pillar view with sparse 2D U-Net in first stage (ablation)RSN · Waymo vehicleL1 BEV AP: 87.7 -> 88.6Improved
use warm starting instead of random sampling of architectures (ablation)LidarNAS search · Waymo vehicleAP and latencyImproved
use perspective view branch (ablation)random architectures · WaymoL1 3D APHurt
use sparse branches on pillar view (ablation)random architectures · Waymolatency (ms): coefficient -0.03 vs dense 3.06Improved
use sparse branches on perspective view (ablation)random architectures · Waymolatency (ms): coefficient 6.90 vs dense -1.65Hurt
use anchor-free head with PointPillars backbonePointPillars-like · Waymo vehicleL1 3D AP: 63.3 -> 69.3Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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