AI Forensics DemoChecking API

Spot deepfakes in seconds.

Upload a face video and run a protected forensic analysis across visual, motion, and consistency signals with supporting evidence.

Upload a videoRun demo examples
Live analysisProtected workflow
Spatial artifacts
Temporal coherence
Global consistency

Output

Verdict + evidence map

Confidence, consistency checks, face-detection quality, and exportable evidence.

5 protected analysisFF++ AUC 1.000Celeb-DF AUC 0.8851Gap reduction 27%CS 668 Capstone · Pace
App repoResearch repoModel weights

Multi-signal forensic review

Visual integrity

Looks for subtle texture, edge, and facial-detail inconsistencies that may indicate manipulation.

Motion consistency

Checks whether facial movement and frame-to-frame transitions remain natural throughout the clip.

Identity coherence

Reviews broader facial consistency so the final assessment is based on more than a single frame.

The workflow checks multiple complementary evidence signals before showing a verdict, so a result is not based on a single visual cue.

Try a pre-loaded example

Preview the video, then run the analysis
real

FF++ C23

FaceForensics++ original

Untouched original from the FF++ in-distribution test split.

fake

FF++ C23

FaceForensics++ Deepfake

FaceSwap/Deepfakes manipulation. The model usually nails this with > 0.99 confidence.

real

Celeb-DF v2

Celeb-DF v2 real

Authentic celebrity footage held out of training entirely.

fake

Celeb-DF v2

Celeb-DF v2 synthesis

Cross-dataset deepfake — harder than FF++ since the manipulation pipeline differs from training.

Or upload your own video

Checking API

Drop a video here

or click to browse · max 50 MB · .mp4 / .mov / .m4v

A preview appears before the file is sent to the GPU backend.