Upload 20 files

Upload 20 files

feat(depth): make Depth Anything V2 Large the default depth estimator

feat(reconstruction): refine camera FOV and constrain Gaussian scale outliers

feat(pointcloud): filter sky/background artifacts in RGB-D backprojection

Upload image_utils.py

Upload image_utils.py

Upload image_utils.py

Upload reconstruction.py

Disbanded SD 1.5 model ID replaced

_handle_hub_error() helper

Space YAML header

Upload depth_estimation.py

Upload app.py

Upload text_to_image.py

Update utils/image_utils.py

Update requirements.txt

Update requirements.txt

Upload requirements.txt

Update requirements.txt

fix: pin jinja2<3.1.0 for gradio 4.x / starlette compatibility

fix: also patch _json_schema_to_python_type for bool schema

fix: patch gradio_client get_type to handle bool schema

fix: patch gradio_client get_type to handle bool schema

fix: change to gradio_client==1.3.0

fix: let HF Spaces manage server binding in demo.launch()

fix: let HF Spaces manage server binding in demo.launch()

fix: pin huggingface_hub<1.0.0 for gradio 4.x compatibility

fix: move __future__ import before audioop patch in app.py

chore: revert gradio to >=4.44.0 and pin pydub==0.25.1

fix: patch audioop at runtime for Python 3.13 / gradio 4.x

fix: upgrade to gradio>=5.0.0 to drop pydub dependency

fix: add pyaudioop for Python 3.13 compatibility

fix: replace open3d with pyntcloud — no py3.13 wheel available

fix: replace open3d with pyntcloud — no py3.13 wheel available

fix: replace open3d with pyntcloud — no py3.13 wheel available

fix: remove obsolete libgl1-mesa-glx from packages.txt

Init

initial commit