Skip to content

Latest commit

 

History

History
57 lines (29 loc) · 2.63 KB

release_note.md

File metadata and controls

57 lines (29 loc) · 2.63 KB

Release Notes

v1.0

2022.12.27

New Features

  • The new version 1.0 of Paddle3D is released, which provides the following features

    • We supports multiple type of 3D perception models, including monocular 3D models SMOKE/CaDDN/DD3D, pointcloud detection models PointPillars/CenterPoint/IA-SSD/PV-RCNN/Voxel R-CNN, BEV visual detection models PETR/PETRv2/BEVFormer, and pointcloud segmentation model SqueezeSegv3

    • We added support for Waymo datasets and now Paddle3D has completed full support for the three open source datasets for autonomous driving

    • Supports automatic mixed-precision training and quantitative deployment capabilities, providing better model acceleration capabilities

    • Supports for sparse convolution, and integrated related SOTA models that are easy to deploy

    • We continue to cooperate with Apollo team to provide one-click deployment of multiple models and integrate them into the perception algorithm part of Apollo to make it easier for developers to debug models

新特性

  • 全部发布Paddle3D 1.0版本,提供了以下特性:

    • 支持多种3D感知模型,包括单目3D模型SMOKE/CaDDN/DD3D,点云检测模型 PointPillars/CenterPoint/IA-SSD/PV-RCNN/Voxel R-CNN,BEV视觉检测模型 PETR/PETRv2/BEVFormer,点云分割模型SqueezeSegv3

    • 新增Waymo数据集支持,完成了对自动驾驶三大开源数据集的全面支持

    • 支持自动混合精度训练以及量化部署能力,提供更好的模型加速能力

    • 新增了对稀疏卷积能力的支持,并集成了稀疏卷积方向的SOTA模型,模型训推一体,便于部署

    • 持续与Apollo进行合作开发,提供多个模型一键部署集成到Apollo的感知算法部分,便于开发者更好地进行模型调试

v0.5

2022.08.09

New Features

  • Release the first version of Paddle3D (that is, v0.5 version), the code of this project is still in beta stage, but under rapid development

  • Release the monocular 3D model SMOKE/CaDDN, the point cloud detection model PointPillars/CenterPoint, and the point cloud segmentation model SqueezeSegv3, all with deployment tutorials

  • Support for KITTI dataset and nuScenes dataset

  • Tutorials for quick access to the Apollo platform are provided for all detection models

新特性

  • 发布Paddle3D 的第一个版本 v0.5,本项目的代码仍在beta阶段,但处于快速迭代中

  • 发布单目3D模型SMOKE/CaDDN,点云检测模型 PointPillars/CenterPoint,以及点云分割模型SqueezeSegv3,均带有部署教程

  • 支持KITTI数据集和nuScenes数据集

  • 为所有的检测模型提供了快速接入Apollo平台的教程