Akash Chowdhury
e9ef5dcea7
- Load Visual Chronometer model with device selection (auto/cpu/cuda) - Predict PhyFPS from video frames using sliding window analysis - Batch prediction with per-segment FPS breakdown - SDPA-optimized attention for better speed and memory efficiency - Auto-downloads model checkpoint from HuggingFace on first use - Progress bar and detailed terminal logging - Input tooltips and node descriptions
11 lines
398 B
Python
11 lines
398 B
Python
"""ComfyUI-PulseOfMotion: Visual Chronometer PhyFPS prediction nodes.
|
|
|
|
Based on "Pulse of Motion" (https://github.com/taco-group/Pulse-of-Motion).
|
|
Predicts the Physical FPS (PhyFPS) of video frames — the true temporal scale
|
|
implied by visual motion dynamics.
|
|
"""
|
|
|
|
from .nodes import NODE_CLASS_MAPPINGS, NODE_DISPLAY_NAME_MAPPINGS
|
|
|
|
__all__ = ["NODE_CLASS_MAPPINGS", "NODE_DISPLAY_NAME_MAPPINGS"]
|