Real-time augmentation for streaming content

By combining lightweight convolutional neural networks and generative adversarial networks, the problem of media content quality degradation in streaming is solved, achieving real-time resolution enhancement and artifact removal on client devices, and is suitable for various encoding and hardware environments.

CN115376035BActive Publication Date: 2026-07-17NVIDIA CORP
View PDF 1 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
NVIDIA CORP
Filing Date
2022-05-13
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

In the process of streaming, existing technologies suffer from quality degradation of media content due to bandwidth and data limitations, resulting in artifacts and reduced resolution. Furthermore, existing methods cannot effectively improve this in real time on client devices.

Method used

Employing lightweight convolutional neural networks and generative adversarial networks, combined with a specific loss function to train the model, it is used for artifact removal and image/video enhancement, suitable for streaming content of different resolutions and bitrates, and can be processed in real time on client devices.

Benefits of technology

It enables real-time resolution enhancement and artifact reduction of media content on client devices, providing a high-definition streaming content experience, and is suitable for various encoding and hardware environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115376035B_ABST
    Figure CN115376035B_ABST
Patent Text Reader

Abstract

The present invention relates to real-time enhancement for streaming content. Real-time content enhancement can be provided using a solution that is light enough to run on a client device, even for high resolution, high bitrate content. The enhancement process can include a neural network that upscales the content to a target resolution while also enhancing the visual quality of the content, for example to sharpen visual aspects of the content and reduce the presence of artifacts. This approach can enable compressed content to be transmitted in a stream over a network in order to save bandwidth and data transfer, while also enabling the content to be upscaled and enhanced in real-time at the client device so that a user or viewer can experience the content at, near, or above their intended or original visual quality.
Need to check novelty before this filing date? Find Prior Art