Contrastive explanations for images with monotonic attribute functions

This invention describes a way for an artificial intelligence to explain its decisions about images. It works by showing what specific visual elements, if added to an image, would make the AI change its mind about what it sees. It can also show which parts of an image are essential for the AI to make its current decision, by removing non-essential parts.

Why it matters: The increasing complexity and widespread deployment of deep learning models since 2021 have amplified the need for robust explainable AI techniques. This method for generating contrastive explanations is now more relevant given the maturity of XAI research and available computational resources.

Status: Not in forceListed as no longer active.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

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