CIA-SSD: Confident IoU-Aware Single-Stage Object Detector From Point Cloud

Abstract

Existing single-stage detectors for locating objects in point clouds often treat object localization and category classification as separate tasks, so the localization accuracy and classification confidence may not well align. To address this issue, we present a new single-stage detector named the Confident IoU-Aware Single-Stage object Detector (CIA-SSD). First, we design the lightweight Spatial-Semantic Feature Aggregation module to adaptively fuse high-level abstract semantic features and low-level spatial features for accurate predictions of bounding boxes and classification confidence. Also, the predicted confidence is further rectified with our designed IoU-aware confidence rectification module to make the confidence more consistent with the localization accuracy. Based on the rectified confidence, we further formulate the Distance-variant IoU-weighted NMS to obtain smoother regressions and avoid redundant predictions. We experiment CIA-SSD on 3D car detection in the KITTI test set and show that it attains top performance in terms of the official ranking metric (moderate AP 80.28%) and above 32 FPS inference speed, outperforming all prior single-stage detectors. The code is available at https://github.com/Vegeta2020/CIA-SSD.

Survey and Systematization of 3D Object Detection Models and Methods

“…The first and more natural solution is to fit a regular grid onto the point cloud, producing a grid cell representation. Many approaches do so by either quantizing point clouds into 3D volumetric grids (Section 6.3.1) (e.g., Song and Xiao, 2014;Zhou and Tuzel, 2018;Shi et al, 2020b) or by discretizing them to (multi view) projections (Section 6.3.2) (e.g., Li et al, 2016;Chen et al, 2017;Beltrán et al, 2018;Zheng et al, 2020).…”

“…Exemplary models using BEV representation can be found in the work from Wang et al (2018), Beltrán et al (2018), Liang et al (2018), Yang et al (2018b), Simon et al (2019b), Zeng et al (2018), , Ali et al (2019), He et al (2020), Zheng et al (2020), Liang et al (2020), and.…”

“…Performancewise, Kuang et al (2020) marily get filled with statistical quantities of the within lying points. Only more recently, first models started to encode the features by deep learning methods (e.g., Lehner et al, 2019;Zheng et al, 2020;Liang et al, 2020).…”

“…To avoid information loss, more recent approaches initially extract features by means of deep learning approaches and then project these features into BEV (e.g., Zheng et al, 2020;Liang et al, 2020).…”

“…For example, Zheng et al (2020) first initialize features of a voxelized point cloud by calculating the mean coordinates and intensities of points in each voxel. Then they apply a sparse convolution and transform the representation to a dense feature map before condensing it on the ground plane to produce a BEV feature map.…”