CIA-SSD: Confident IoU-Aware Single-Stage Object Detector From Point Cloud
Wu Zheng, Weiliang Tang, Sijin Chen, Li Jiang, Chi-Wing Fu

TL;DR
CIA-SSD is a novel single-stage point cloud detector that improves localization and classification confidence alignment through a new confidence rectification and IoU-weighted NMS, achieving top performance and high speed.
Contribution
The paper introduces CIA-SSD, a lightweight and confident IoU-aware detector with novel modules for feature fusion and confidence rectification, enhancing accuracy and efficiency.
Findings
Achieves 80.28% moderate AP on KITTI dataset.
Runs at over 32 FPS, outperforming prior single-stage detectors.
Outperforms existing methods in accuracy and speed.
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…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
Taxonomy
TopicsAdvanced Neural Network Applications · Video Surveillance and Tracking Methods · Automated Road and Building Extraction
