Geen Opslag Crisisbescherming: Geüploade bewijzen worden uitsluitend in het vluchtige geheugen geanalyseerd en direct gewist. Geen logs, geen back-up.
CogVideoX-5B Video Detector: Zhipu AI Generation Forensics
Analyzing 3D variational autoencoder video diffusion models for synthetic motion and spatial smoothing.
Authenticiteit verifiëren met Sealed Rose
Directe Neurale Analyse · Gratis · Geen Account Vereist · Geen Opslag · Direct in Geheugen
CogVideoX-5B (developed by Zhipu AI) uses an expert 3D VAE to compress spatial and temporal dimensions. Sealed Rose inspects 3D VAE reconstruction bounds to identify synthetic video generated by this architecture.
Kritieke Waarschuwingssignalen
1Soft Focus on Small Background Objects
Background pedestrians or distant signage exhibit a melted, dreamlike texture.
3D VAE Compression Signatures
- Temporal Latent Blurring: 3D temporal convolutions compress motion across 4 frames, causing micro-ghosting in fast movements.
Hoe uzelf te beschermen en handelen
Run Video Through Sealed Rose
Detect 3D VAE compression patterns.
Veelgestelde Vragen
What is CogVideoX-5B?▼
It is an open-weights video generation model capable of generating up to 6-second clips at 720p resolution.
Beschikbaar in 23 Talen: