ComfyUI Node
Deep Exemplar Video Colorization (Original)
A ComfyUI node in DeepExemplar/Video with 8 inputs and 2 outputs.
Deep Exemplar Video Colorization (Original)
- video_frames
- reference_image
- colorized_frames
- performance_report
◄frame_propagatetrue►
◄use_half_resolutiontrue►
◄target_width768►
◄target_height432►
◄use_torch_compilefalse►
◄use_sage_attentionfalse►
CategoryDeepExemplar/Video
Inputs (8)
| Name | Type | Default | Description |
|---|---|---|---|
| video_frames | IMAGE | Batch of video frames to be colorized | |
| reference_image | IMAGE | Color reference image that provides the color palette | |
| frame_propagate | BOOLEAN | true | Enable temporal propagation: use previous frame's colorization to inform current frame (improves consistency) |
| use_half_resolution | BOOLEAN | true | Process at half resolution for speed, then upscale (2x faster with minimal quality loss) |
| target_width | INT | 76816–4096 | Output width (will be adjusted to nearest multiple of 32) |
| target_height | INT | 43216–4096 | Output height (will be adjusted to nearest multiple of 32) |
| use_torch_compile | BOOLEAN | false | Enable torch.compile optimization for 10-25% speedup (may increase first-run compilation time) |
| use_sage_attention | BOOLEAN | false | Enable SageAttention for faster attention computation (requires sageattention package) |
Outputs (2)
| Name | Type | Description |
|---|---|---|
| colorized_frames | IMAGE | — |
| performance_report | STRING | — |