AnyCamVLA-LVSM

Tools

• anycamvla_lvsm_demo_synthesize_novel_view Synthesize a novel view between two source views of the same scene. LVSM reconstructs the target view by triangulating the two posed source views, so it needs two GENUINELY different camera viewpoints of the same scene (not a duplicated image) together with their relative camera pose. The bundled examples carry the real camera poses of each pair; the target camera is placed by interpolating between the two source cameras (``view_mix``) and then nudged by the small translation / rotation offsets. If a user passes a file as an input, use the upload_file_to_gradio tool, if present, to upload the file to the gradio app and create a Gradio File Input. Then use the returned path as the input to the tool

No tool matches.

As the server's own listing gives them. We never call a tool.