Repository files navigation

PytorchAutoDrive is apure Python framework includes semantic segmentation models, lane detection models based onPyTorch. Here we provide full stack supports from research (model training, testing, fair benchmarking by simply writing configs) to application (visualization, model deployment).

Paper:Rethinking Efficient Lane Detection via Curve Modeling (CVPR 2022)

Poster:PytorchAutoDrive: Toolkit & Fair Benchmark for Autonomous Driving Research (PyTorch Developer Day 2021)

This repository is under active development, results with models uploaded are stable. For legacy code users, please checkdeprecations for changes.

A demo video from ERFNet:

Various methods on a wide range of backbones,config based implementations,modulated andeasily understood codes, image/keypoint loading, transformations andvisualizations,mixed precision training, tensorboard logging anddeployment support with ONNX and TensorRT.

Models from this repo are faster to train (single card trainable) and often have better performance than other implementations, seewiki for reasons and technical specification of models.

* The UDA baseline setup, with Cityscapesval set as validation.

We provide solid results (average/best/detailed), training time, shell scripts and trained models available for download inMODEL_ZOO.md.

Please prepare the environment and code withINSTALL.md. Then follow the instructions inDATASET.md to set up datasets.

Get started withLANEDETECTION.md for lane detection.

Get started withSEGMENTATION.md for semantic segmentation.

Refer toVISUALIZATION.md for a visualization & inference tutorial, for image and video inputs.

Refer toBENCHMARK.md for a benchmarking tutorial, including FPS test, FLOPs & memory count for each supported model.

Refer toDEPLOY.md for ONNX and TensorRT deployment supports.

CheckoutADVANCED_TUTORIAL.md for advanced use cases and how to code in PytorchAutoDrive.

Refer toCONTRIBUTING.md for contribution guides.

If you feel this framework substantially helped your research or you want a reference when using our results, please cite the following paper that made the official release of PytorchAutoDrive:

@inproceedings{feng2022rethinking,  title={Rethinking efficient lane detection via curve modeling},  author={Feng, Zhengyang and Guo, Shaohua and Tan, Xin and Xu, Ke and Wang, Min and Ma, Lizhuang},  booktitle={Computer Vision and Pattern Recognition},  year={2022}}

PytorchAutoDrive is maintained by Zhengyang Feng (voldemortX) and Shaohua Guo (cedricgsh).

Contributors (GitHub ID):kalkun,LittleJohnKhan,francis0407,PannenetsF,bjzhb666

People who sponsored us (e.g., with hardware):Lizhuang Ma,Xin Tan, Junshu Tang (junshutang), Fengqi Liu (FengqiLiu1221)