ComfyUI Node
CoachBate H3 Subject Count
Runs a YOLO detector over sampled frames and returns the number of subjects the clip shows - the largest per-frame count that holds on at least min_fraction of the sampled frames, capped by max_subjects. Wire to Pixaroma Loop Start 'total' so a per-subject pass (select_index = loop index) runs once for each.
CoachBate H3 Subject Count
- images
- subjects
- report
◄detector▾►
◄confidence0.30►
◄sample_every1►
◄min_fraction0.05►
◄max_subjects20►
CategoryCoachBate/H3
Inputs (6)
| Name | Type | Default | Description |
|---|---|---|---|
| images | IMAGE | — | |
| detector | COMBO | The same detector the tracker uses. | |
| confidence | FLOAT | 0.300.05–0.95 | — |
| sample_every | INT | 11–60 | Detect on every nth frame. 1 = every frame. |
| min_fraction | FLOAT | 0.050.01–1 | A count must appear on at least this fraction of the sampled frames to be believed, so a single-frame false positive does not add a whole extra render pass. |
| max_subjects | INT | 201–50 | Cap. Each subject is a full H3 render of the clip. 1 forces a single pass. |
Outputs (2)
| Name | Type | Description |
|---|---|---|
| subjects | INT | — |
| report | STRING | — |