Replace install_croco function with a direct pip installation command for the curope model dependency in app.py, streamlining the installation process and removing the previous compilation logic.

Enhance install_croco function in app.py by adding checks for source file existence, improved CUDA architecture handling, and a cleanup process for the build directory. Implement an alternative compilation method using setup.py if the primary method fails, ensuring robust error handling and better user feedback during the installation of the curope CUDA extension.

Refactor install_croco function in app.py to compile the curope CUDA extension using torch.utils.cpp_extension.load() instead of subprocess calls, improving compatibility with zerogpu. Added error handling and fallback to PyTorch implementation if compilation fails.

Remove redundant import of torch in app.py and add it back for proper functionality with GPU context and autocasting.

Refactor app.py to encapsulate the automatic installation of the curpo model dependency within a dedicated function, utilizing GPU context and autocasting for improved performance.

Implement visualization directory management in app.py and navigation.py. Added clear_visualization_directory function to prevent users from seeing each other's generated images. Updated IMAGE_PATHS with new sample images and removed unused imports.

Update links in app.py for the VMem paper to point to the correct URL.

Update NAVIGATION_FPS to 13 and correct links in app.py for the VMem paper and GitHub repository.

Remove unused image 'living_room_2.jpeg' from IMAGE_PATHS and update corresponding labels in app.py for clarity.

Enhance app.py UI with new logo and updated title; add demo assets to .gitattributes. Remove unused main function from navigation.py and streamline .gitignore.

Update app.py and navigation.py for UI enhancements and image path correction. Changed primary color theme in app.py and updated image path in navigation.py to reflect new asset.

Refactor inference configuration and pipeline logic; removed unused parameters and improved frame selection process. Updated inference settings in inference.yaml and streamlined surfel model initialization in pipeline.py.

Update image assets and refactor app.py for improved functionality. Added new images, removed outdated ones, and updated paths in IMAGE_PATHS. Renamed render_demo3 function to render_demonstrate for clarity. Adjusted UI styles and fixed image loading issues.

Refactor app.py to include automatic installation of curpo model dependency and remove redundant subprocess import

Add automatic installation of curpo model dependency in app.py and comment out GitHub link in requirements.txt

Update requirements.txt to use GitHub repository for curpo model dependency

Fix path for CUT3R model directory in requirements.txt

Update requirements.txt to correct path for CUT3R model directory

Update project configuration and dependencies; modify .gitignore, adjust README title, and refine inference settings

Update README with project details and remove training script

Add initial project structure with core files, configurations, and sample images

initial commit