ComfyUI Node
Frame Description Aggregator (no captioning)
Joins per-frame descriptions you type (one line per selected frame) into a narrative and key-moment JSON, using histogram scene boundaries from the frames. It does not caption images and runs no model.
Frame Description Aggregator (no captioning)
- video_frames
- narrative
- key_moments
◄frame_interval1►
◄frame_descriptions►
◄boundary_threshold0.35►
CategoryParametric/Video
Inputs (4)
| Name | Type | Default | Description |
|---|---|---|---|
| video_frames | IMAGE | Video frames. | |
| frame_interval | INT | 11–256 | Frame interval. From 1 to 256. Default 1. |
| frame_descriptions | STRING | Frame descriptions. | |
| boundary_threshold | FLOAT | 0.350–1 | Boundary threshold. From 0 to 1. Default 0.35. |
Outputs (2)
| Name | Type | Description |
|---|---|---|
| narrative | STRING | — |
| key_moments | STRING | — |