Conversational AI-encoded language for data compression

This invention describes a way to compress video by analyzing a person's behavior, specifically their facial expressions, throughout the video. Instead of storing the full video, the system saves a single picture of the person and a sequence of codes representing their facial expressions over time. When you want to watch the video, a special program uses the picture and the expression codes to recreate the video. The claims focus specifically on detecting and encoding facial expressions.

Why it matters: Filed in 2022, the rapid advancements in generative AI and real-time facial analysis since then could significantly improve the accuracy of behavior recognition and the realism of video synthesis, making the reconstruction more convincing and efficient.

Status: ExpiredProtection ended and was never renewed
How hard to build: SpecializedAI/ML models, video synthesis, real-time processing

3 AI reinvention directions for this patent

See all three

AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.

Reinvent this with AI

Where this one came from

Every morning we go through millions of abandoned patents and pick out 15 worth your time. This was one of today’s.

This one

+10