bhack
bhack
Probably a light segmentation component it could help in the model to understand what we are tracking with the grid allocation in cluster tracks.
@nikitakaraevv probably you could explore any recent active learning approach to select new synthetic or real useful data samples.
@nikitakaraevv About 2. Do you think we could adapt some techniques used in VO like https://chiaki530.github.io/projects/leapvo/
See section 4.4 of the related paper.
Any progress on this for v2?
Have you re-evaluated this recently?
Ok same issue with the sfm just recently pointed in the Readme. It starts to seem the new/default policy of Meta research teams.
Just the traditional ping after every new release.. Do you plan to release v3 under an Opensource/OSI license?
```cmake -- Could NOT find CUDNN (missing: CUDNN_LIBRARY_PATH CUDNN_INCLUDE_PATH) CMake Warning at cmake/public/cuda.cmake:250 (message): Cannot find cuDNN library. Turning the option off Call Stack (most recent call first): cmake/Dependencies.cmake:44 (include)...
Can we really start to regularly test the nightly build in CI using exact the same instruction we have in: https://github.com/pytorch/pytorch?tab=readme-ov-file#from-source https://github.com/pytorch/pytorch/blob/main/CONTRIBUTING.md