LPM 1.0 breakthrough enables realistic 45-minute talking head videos from a single photo with real-time processing

LPM 1.0 converts a single image into a talking character with synchronized lip movements

The model generates up to 45 minutes of continuous video while running in real time

Generated videos include facial expressions and emotional reactions beyond basic lip-syncing

Currently remains in research phase and has not been released for public use