公开(公告)号：US20250086828A1

公开(公告)日：2025-03-13

申请号：US18818790

申请日：2024-08-29

Applicant: TuSimple, Inc.

Inventor： Dongqiangzi YE ,  Yufei XIE ,  Weijia CHEN ,  Zixiang ZHOU ,  Lingting GE

IPC: G06T7/73 ,  G01S17/89 ,  G06T7/12 ,  G06T17/00 ,  G06V10/25 ,  G06V10/80

Abstract: An image processing method includes performing, using images obtained from one or more sensors onboard a vehicle, a 2-dimensional (2D) feature extraction; performing, a 3-dimensional (3D) feature extraction on the images; detecting objects in the images by fusing detection results from the 2D feature extraction and the 3D feature extraction.

公开(公告)号：US20250086802A1

公开(公告)日：2025-03-13

申请号：US18434501

申请日：2024-02-06

Applicant: TuSimple, Inc.

Inventor： Dongqiangzi YE ,  Zixiang ZHOU ,  Weijia CHEN ,  Yufei XIE ,  Yu WANG ,  Panqu WANG ,  Lingting GE

IPC: G06T7/10 ,  G06T9/00 ,  G06T17/00

Abstract: A method of processing point cloud information includes converting points in a point cloud obtained from a lidar sensor into a voxel grid, generating, from the voxel grid, sparse voxel features by applying a multi-layer perceptron and one or more max pooling layers that reduce dimension of input data; applying a cascade of an encoder that performs a N-stage sparse-to-dense feature operation, a global context pooling (GCP) module, and an M-stage decoder that performs a dense-to-sparse feature generation operation. The GCP module bridges an output of a last stage of the N-stages with an input of a first stage of the M-stages, where N and M are positive integers. The GCP module comprises a multi-scale feature extractor; and performing one or more perception operations on an output of the M-stage decoder and/or an output of the GCP module.