概要説明
マルチカメラの3D
既存技術より優れ
キー技術
-
Image-view Encoder
-
View Transformer
マルチカメラの3D
Image-view Encoder
View Transformer
Categorical Depth Distribution Network for Monocular 3D Object Detection
Cody Reading Ali Harakeh Julia Chae Steven L. Waslander
University of Toronto Robotics Institute
https://arxiv.org/pdf/2103.01100