zhang.haojie
zhang.haojie
Hello, I'm currently working with the `GAP` or `LAP` and would appreciate some guidance on the shape requirements for the inference interface parameters. For example `mask_pred`, `image`, `img_mst_tree`. Could you...
I noticed that your validation metric for [hallo](https://github.com/fudan-generative-vision/hallo) on the HDTF dataset is 501, while the original metric is 173. I would like to understand the specific details of the...
If I don't want to compress in the temporal dimension when using VAE compression, can I still directly use CogVideoX? What are some recommended methods? Thank you!