daohu527
commented
2 months ago I will create a new plan and organize some discussions, welcome to participate!
Yes, I feel some inconveniences with the current use of autonomous driving, so I want to make an autonomous driving system that is simple, easy to use, and has rich algorithms and interfaces.
I have just started this project. Of course, I actually have a lot of ideas. The most urgent thing at the moment is to perfect the perception part first.
I currently use pytorch to implement several functions such as object detection, segmentation, and posture detection. I hope that what you see is what you get.
input
For example you can specify the input source as:
- Files: txt (include path), pictures, videos, point clouds, bags
- Network files: URL
- Stream: video stream
- Screenshot: Screenshot
- Messages: subscribe topics
output
After model inference, the visual interface can be directly output.
- Visualized results with labels
- Inference results save desired format
I hope it can implement various algorithm libraries as elegantly as possible. There are already some ideas for these, and I hope it is easy to simulate.
github-actions[bot]
Hi @daohu527
I’m very interested in contributing to this project and I’ve wanted a python ADS for a long time. How should we contribute to this project?