1. Add 1 optional IMAGES input port named a0_image.
2. Add 2 new bitmap stitching modes: equal-width vertical stitching (top to bottom), equal-height horizontal stitching (left to right).
3. Add the function of displaying image filenames on the stitched tiles, with customizable text colors (16 colors).
4. Optimized the tile saving mode, including three options: none/save single block/save single image.
5. Simply connect to the b6 port to view the usage help for this node.
6. Standardized the port naming on the node page.
Updated description to include new features in V1.1.
- ✨ Added 2 new bitmap stitching modes: equal-width vertical stitching (top to bottom), equal-height horizontal stitching (left to right)
- ✨ Optimized the tile saving mode, including three options: none/save single block/save single image
- ✨ Added the function of displaying image filenames on the stitched tiles, with customizable text colors (16 colors)
- ✨ Simply connect to the b6 port to view the usage help for this node
- 🎨 Standardized the port naming on the node page
Halo-Lipsy is a native Wav2Lip implementation for ComfyUI that finally works
on AMD APUs with unified memory (Strix Halo, etc.).
Key features:
- No subprocess hacks - runs natively in ComfyUI process
- Face detection on CPU to prevent memory conflicts
- Safe tensor casting for unified memory compatibility
- Sync tuning and edge blending options
- Works on ROCm, AMD APUs, and NVIDIA
Tested on Ryzen AI Max+ 395 (Strix Halo) with 64GB unified memory and ROCm 7.11.
Repository: https://github.com/bkpaine1/Halo-Lipsy
Co-authored-by: bkpaine1 <bkpaine1@users.noreply.github.com>
