Closed PINTO0309 closed 9 months ago
How well does it work with pointing your head backwards? Released model is for 300W-LP, which does not contain a wide range of motion, which was indicated by the paper
I do not measure MAE because it is too cumbersome. Better than WHENet in sensory evaluation. Pitch is very weak. Well, maybe the Pitch inference just doesn't work well when the range of head detection is large.
https://github.com/thohemp/6DRepNet360/assets/33194443/6587bcb1-6a09-4cba-9830-af4bf0e89a24
If you are interested, I have added end to end export to onnx, with euler angles baked into onnx, tested with trt
Thanks.
However, I think you should open an issue on my repository or open another issue in this repository, since the content is different from the main topic of this issue.
This issue asks, "What is the license for the model?"
Hello @PINTO0309,
thanks for your work. I add the license to the repo.
Thank you for sharing your great code. 😺 I have enjoyed it very much since the 6DRepNet generation. Thank you.
What is the license for this model? I'd like to cite it to the repository I'm working on if possible, but I want to post the license correctly.
https://github.com/PINTO0309/PINTO_model_zoo/tree/main/423_6DRepNet360
My test results - Gold-YOLO (onnx) + 6DRepNet360 (50 batch / onnx)
https://github.com/thohemp/6DRepNet360/assets/33194443/e2fa511a-68f8-4b9e-80d9-87141b6285cc