USPTO Patent US12586350B2: Adobe Inc. AI for audio/video representations

ChangeBridge: Patent Grants - AI & Computing (G06N)

Published March 24th, 2026

Detected March 25th, 2026

Summary

The USPTO has granted patent US12586350B2 to Adobe Inc. for a method of determining audio and video representations using self-supervised learning. The patent details training machine learning models to learn short-term, long-term, and semantic features of input data.

View original document View source feed page

What changed

The United States Patent and Trademark Office (USPTO) has granted patent US12586350B2 to Adobe Inc. The patent, titled 'Determining audio and video representations using self-supervised learning,' describes a method for training systems to generate audio and video representations. This involves using machine learning models trained with contrastive and temporal learning tasks to analyze video signals, thereby learning short-term, long-term, and semantic features of the data.

This patent grant is primarily an intellectual property matter and does not impose new regulatory obligations on companies. However, it signifies a technological advancement in AI for media representation, which may influence future product development and competitive landscapes in the technology sector. Companies developing AI for audio and video processing should be aware of this patent's scope.

Source document (simplified)

← USPTO Patent Grants

Determining audio and video representations using self-supervised learning

Grant US12586350B2 Kind: B2 Mar 24, 2026

Assignee

Adobe Inc.

Inventors

Simon Jenni, John Collomosse

Abstract

Embodiments are disclosed for training a system to generate audio and video representations using self-supervised learning. The method may include receiving a video signal including an audio component and a video component. A first machine learning model is trained to determine a representation of the audio component using a contrastive learning task and a temporal learning task. A second machine learning model to determine a representation of the video component using the contrastive learning task and the temporal learning task. By training the machine learning models using both contrastive learning tasks and temporal learning tasks, the machine learning models learn short term features, long term features, and semantic features of input data.