ComfyUI Node
Mask Mix (holes + semantic)
A ComfyUI node in SplatKit with 6 inputs and 3 outputs.
Mask Mix (holes + semantic)
- holes
- semantic
- mask
- preview
- composite_mask
◄semantic_weight0.50►
◄semantic_threshold0.50►
◄semantic_grow_px0►
◄hole_grow_px0►
CategorySplatKit
Inputs (6)
| Name | Type | Default | Description |
|---|---|---|---|
| holes | MASK | The disocclusion mask, 1 = hole. From HiRes `hole_mask` inverted (hole_mask is 1 = KEPT). | |
| semanticopt | IMAGE | The semantic region rendered through the SAME rig -- run a second HiRes Pano Fly-Through with identical settings and the SAM3 mask wired into `texture_override`. | |
| semantic_weightopt | FLOAT | 0.500–1 | How much of these regions the video model reinvents. 0 = untouched, 1 = fully regenerated (identity lost). 0.4-0.6 buys parallax and moving reflections while the plant stays the same plant. |
| semantic_thresholdopt | FLOAT | 0.500–1 | Cutoff applied to the rasterised semantic mask before weighting. Raise if the region bleeds past its edges. |
| semantic_grow_pxopt | INT | 00–128 | Dilate the semantic region. A few px lets the model move a leaf silhouette instead of being pinned to its outline. |
| hole_grow_pxopt | INT | 00–128 | Dilate the holes. Leave at 0 if a Grow Mask already ran. |
Outputs (3)
| Name | Type | Description |
|---|---|---|
| mask | MASK | Weighted mask: 1 in the holes, `semantic_weight` on the plants/windows. This is what drives masked video conditioning, or the green/black fill of a control video. |
| preview | IMAGE | The same thing as a viewable IMAGE. |
| composite_mask | MASK | Binary stencil for the final Image Composite Masked: 1 wherever the model was allowed to change anything. Use THIS there, never the weighted mask -- compositing a regenerated plant back at 0.5 ghosts it against the original. |