Ashish Patel 🇮🇳’s Post

𝗗𝗮𝘆-𝟰𝟲𝟰 𝗖𝗼𝗺𝗽𝘂𝘁𝗲𝗿 𝗩𝗶𝘀𝗶𝗼𝗻 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 RBGNet: Ray-based Grouping for 3D Object Detection by Center for Data Science, Peking University Follow me for a similar post: Ashish Patel ------------------------------------------------------------------- 𝗜𝗻𝘁𝗲𝗿𝗲𝘀𝘁𝗶𝗻𝗴 𝗙𝗮𝗰𝘁𝘀 : 🔸 This paper is published CVPR2022. 🔸 Github: https://lnkd.in/gG7rHFqB ------------------------------------------------------------------- 𝗜𝗠𝗣𝗢𝗥𝗧𝗔𝗡𝗖𝗘 ➡️ As a fundamental problem in computer vision, 3D object detection is experiencing rapid growth. ➡️ To extract the point-wise features from the irregularly and sparsely distributed points, previous methods usually take a feature grouping module to aggregate the point features to an object candidate. ➡️ However, these methods have not yet leveraged the surface geometry of foreground objects to enhance grouping and 3D box generation. ➡️ In this paper, we propose the RBGNet framework, a voting-based 3D detector for accurate 3D object detection from point clouds. ➡️ In order to learn better representations of object shape to enhance cluster features for predicting 3D boxes, we propose a ray-based feature grouping module, which aggregates the point-wise features on object surfaces using a group of determined rays uniformly emitted from cluster centers. ➡️ Considering the fact that foreground points are more meaningful for box estimation, we design a novel foreground biased sampling strategy in downsample process to sample more points on object surfaces and further boost the detection performance. ➡️ Our model achieves state-of-the-art 3D detection performance on ScanNet V2 and SUN RGB-D with remarkable performance gains. #computervision #artificialintelligence #technology