Intelligent control with hierarchical stacked neural networks

This invention describes a system that uses multiple layers of neural networks to analyze various types of data. While the abstract focuses on identifying unusual or "aberrant" patterns in text messages by looking at their words and grammar, the actual claims broaden this to analyzing any data pattern, specifically mentioning images and video. The system aims to quantify how much a data pattern deviates from what's expected.

Why it matters: Since 2014, deep learning architectures like Transformers have revolutionized the ability of neural networks to understand and detect subtle anomalies in complex data, whether text, images, or video. This makes the core concept of quantifying deviations in data patterns far more powerful and practical to implement today.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedDeep learning software, data 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