Bump version to 0.11.2 and update uv version to 0.11.3
- Updated version in pyproject.toml and blender_manifest.toml to 0.11.2. - Enhanced uv executable path documentation in client.py. - Added inference_mode parameter to reconstruct_request function in client.py. - Implemented uv download functionality in install_env.py with improved error handling and checksum verification. - Modified install_uv function to download and extract uv binary instead of using bundled version. - Updated job progress handling in operators.py to track progress stages. - Improved download progress reporting in worker.py and adjusted related tests. - Refactored tests to accommodate changes in status reporting and model downloading. - Removed vendor.lock.toml as part of the cleanup process.
This commit is contained in:
@@ -22,6 +22,8 @@
|
||||
因此材质直接用 `.foreground` 同时提供颜色和 Mask,不再额外加载原图 Image Texture。
|
||||
Image Empty 引用视频或序列帧时,DA3 与 MoGe-2 会按 Empty 的 Image Frames 数值逐帧估计
|
||||
深度(未设置时为完整帧数,显式设为 1 时只处理 1 帧),起始帧跟随 Empty 的 Offset,与背景蒙版范围一致;
|
||||
DA3 会复用同一个 ONNX Session 对每帧独立推理,再合并为原有序列输出,避免多帧 Attention 的平方级显存占用。
|
||||
该行为只用于 Depth Plane 与 Surface 转换;Debug 多帧输出仍联合推理。
|
||||
Offset 超出视频长度时同样把窗口收进视频内。材质节点与置换贴图的 Start 保持与原 Empty
|
||||
一致;Original 模式的视频或序列帧使用 `frame.0001.jpg` 等原图序列作为颜色与深度,
|
||||
`foreground.0001.png` 只提供前景 Alpha。
|
||||
|
||||
Reference in New Issue
Block a user