Yi Yang

Hi @weirenorweiren , we are so far not very familiar with windows issues. From what I can tell, if just for live_demo.py, you don't need tensorflow dependency. For jaxlib, our...

Hi @weirenorweiren , we just released a [pytorch live demo](https://github.com/google-deepmind/tapnet/blob/main/tapnet/pytorch_live_demo.py) so jaxlib dependency is also not necessary any more. Feel free to give it a try.

For tracking points from another query frame, you can extract the query features from your query frame, then run the OnlineTAPIR inference pipeline. i.e. query_features = online_model_init(frames[None, query_frame_index:query_frame_index + 1],...

Hi @PinxueGuo We just update our [colab notebooks](https://colab.sandbox.google.com/github/deepmind/tapnet/blob/master/colabs/torch_causal_tapir_demo.ipynb) to allow select points in the middle frames in Online TAPIR. Please give it try at your convinience. Note that this is...

Hi @weirenorweiren , upon your request, we have added the [pytorch live demo]( https://github.com/google-deepmind/tapnet/blob/main/tapnet/pytorch_live_demo.py) to our codebase.

Hi @weirenorweiren , we have not conduct any performance test for the pytorch live demo, presumably the runtime shall be relatively the same under the same hardware. Feel free to...

Just did a quick fix (or removal due to refactor), please check if the links are working now. Thank you for reporting the issues!

We are investigating this issue now. Thanks for your "patience". Apologize for the long waiting time.

@gorkaydemir We just fix the online model performance gap for the PyTorch Causal BootsTAPIR. Please give it a retry: https://colab.sandbox.google.com/github/deepmind/tapnet/blob/master/colabs/torch_causal_tapir_demo.ipynb Thanks again for reporting the issue.

I will close it now since we are not planning to implement pytorch training for TAPIR so far. We may re-open if there is indeed a version. We also welcome...