AI Limitations & Best Practices
Known Limitations
- • Occasional flickering in fast-moving scenes
- • Motion blur may affect transformation quality
- • Complex backgrounds might show artifacts
- • Side profiles may not transform as accurately as front faces
- • Very dark or overexposed videos may have reduced quality
Recommended Input Quality
- • Clear portrait photo, ideally 720px or wider
- • Good lighting conditions
- • Clear, visible faces
- • JPG, PNG, or WebP format
- • One main face works better than crowded group shots
Tips for Best Results
- • Use videos with front-facing subjects
- • Ensure faces are at least 10% of the frame
- • Avoid extreme angles or rotations
- • Keep videos under 30 seconds for optimal processing
- • Use the preview feature before finalizing
- • Try different strength settings (0.3-0.5 recommended)
Failure Cases & Solutions
No faces detected
Ensure faces are clearly visible and at least 64×64 pixels
Processing timeout
Try reducing video length or resolution
Poor quality output
Check input video quality and lighting conditions