Upload Video
Drop Video Here
- or -
Click to Upload
🎞️ Example Videos — (click to expand)
▼
⚡ Model Variant
Radio
Fast (b0): 8.7M params | Pro (b5): 50.3M params
fast
pro
🌍 Training Domain
Radio
🇰🇷 KoDF — East Asian faces
🌟 Celeb-DF-v2 — Celebrity faces
🌍 FF++ — Western faces
🔍 Analyze
Result
Class Probabilities
Model Guide
⚡ Fast (b0)
🔥 Pro (b5)
Params
8.7M
50.3M
Resolution
224×224
384×384
Speed
Faster
Slower
Accuracy
High
Higher
Domain Guide
Domain
Training Data
Best For
KoDF
62K real / 175K fake (Korean)
East Asian faces
Celeb-DF-v2
890 real / 5.6K fake (Celebrities)
Celebrity faces
FF++
1K real / 6K fake (4 manipulation types)
Western faces
⚙️ Advanced Settings
▼
Adjust inference parameters. Defaults are recommended for most cases.
Num Frames
Number of frames sampled from the video
↺
5
20
Margin Ratio
Margin around the detected face bounding box
↺
0
0.5
TTA Horizontal Flip
⚠️ > 0.0 doubles inference time per frame
↺
0
1
Aggregation Mode
conf: heuristic (recommended) · mean: average · vote: majority
conf
mean
vote
Face Detection Confidence
YOLO face detection confidence threshold
↺
0.1
0.9
Min Face Ratio
Minimum face-to-frame area ratio to process
↺
0.001
0.1
📊 Frame-level Results
▼