revert

refactor: enhance track visualization functionality

fix

test

test

move

add ftfy==6.3.1 to requirements

move wan pipeline to cuda in zero gpu inference time

load wan in global

revert

fix

fix

fix

fix

fix

add ttm pipelines

fix

fix

add wan ttm

revert

add some logs to check run issue

change req to rebuild

Revert "init ttm"

fix imports

test another version

fix imports

refactor code

init ttm

Add GPU-accelerated spatial tracking function and refactor video processing flow

Refactor video processing functions to streamline TTM input generation; add option for generating motion signal and mask videos, and update UI components accordingly.

Add motion signal and mask video generation to TTM processing; enhance UI with speed control and output mode selection

Improve bundle adjustment configuration by ensuring constant camera pose is set only for valid registered images.

prevents the "Check failed: HasImage(image_id)" error by ensuring we only call set_constant_cam_pose() with a valid image_id that actually exists in the reconstruction.

fix utils3d version

fix utils3d version

fix utils3d.torch.utils.depth_edge function path

Add Gradio UI demo for video-to-pointcloud renderer with camera movement selection

fix pycare package version

remove some package

add packages.txt and downgrade pycolmap and pyceres packages in requirements.txt

cite our paper (#6)

add online

support HubMixin

support HubMixin

renew_files

Upload 3 files (#5)

Update app.py (#4)

Upload 3 files (#3)

add readme

Cleaned history, reset to current state