fyang064

The 3dmatch_sample in the example_data is a good example to learn and modify. However, I met some troubles when I tried to generate my own data using the 3DSmooth. ...

As mentioned in the report "Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasets," 8 80GB A100 GPUs were used for training.

> 可以不用输入disp，disp本身是不参与配准过程的，我只是为了可视化扭曲图像才使用它的. I'm wondering if you could also provide the npy file like default='../dataset/raw/ctest/Road/disp', which seems like a required argument in the test_reg.py.

> Since infini attention uses segmentation which is mainly focused on reducing memory usage and computational cost into O(N) so if you use very long seq such as seq len...