Frida H

A 1TB video dataset. Our estimated time for all the footage would be between 100-250 hours. This includes raw, high-definition (H.264) footage from real movie productions. Includes a diverse range of cinematic scenes—from 20-second takes to multi-minute sequences—with natural lighting, on-set audio, and varied actor performances. Ideal for training AI models in scene understanding, emotion recognition, video editing automation, and cinematic analysis.