18272862. Video Coding With Guided Machine Learning Restoration simplified abstract (Google LLC)
Contents
Video Coding With Guided Machine Learning Restoration
Organization Name
Inventor(s)
Urvang Joshi of Mountain VIew CA (US)
Sarah Parker of San Francisco CA (US)
Elliott Karpilovsky of Santa Clara CA (US)
Debargha Mukherjee of Cupertino CA (US)
Video Coding With Guided Machine Learning Restoration - A simplified explanation of the abstract
This abstract first appeared for US patent application 18272862 titled 'Video Coding With Guided Machine Learning Restoration
Simplified Explanation
The patent application describes image coding using guided machine learning restoration, where a restored frame is obtained by decoding a reconstructed frame and then restoring it using a trained guided convolutional neural network. The restored block is generated by inputting the reconstructed block into the neural network and obtaining a guided neural network prediction based on guide parameter values and neural network output channel predictions.
- Decoding a reconstructed frame to obtain frame data
- Restoring the frame using a trained guided convolutional neural network
- Generating a restored block by inputting the reconstructed block into the neural network
- Obtaining a guided neural network prediction based on guide parameter values and neural network output channel predictions
- Outputting the restored frame
---
- Potential Applications
This technology could be applied in video compression, image restoration, and video streaming services.
- Problems Solved
This technology helps in improving the quality of compressed images and videos, reducing artifacts, and enhancing visual content.
- Benefits
The benefits of this technology include higher image quality, improved compression efficiency, and better visual experience for users.
- Potential Commercial Applications
Commercial applications of this technology could include video streaming platforms, surveillance systems, and medical imaging devices.
- Possible Prior Art
One possible prior art could be traditional image and video compression techniques that do not utilize guided machine learning for restoration.
---
- Unanswered Questions
- How does this technology compare to existing image restoration methods?
This article does not provide a direct comparison with traditional image restoration methods or other machine learning-based approaches.
- What are the limitations of using guided machine learning for image coding?
The article does not discuss any potential limitations or challenges associated with implementing guided machine learning for image coding.
Original Abstract Submitted
Image coding using guided machine learning restoration may include obtaining reconstructed frame data by decoding, obtaining a restored frame by restoring the reconstructed frame, and outputting the restored frame. Obtaining the restored frame may include obtaining a reconstructed block, obtaining guide parameter values, obtaining a restored block, and including the restored block in the restored frame. Obtaining the restored block may include inputting the reconstructed block to an input layer of a trained guided convolutional neural network, wherein the neural network is constrained such that an output layer has a defined cardinality of channels, obtaining, from the output layer, neural network output channel predictions, obtaining a guided neural network prediction as a linear combination of the guide parameter values and the neural network output channel predictions, and generating the restored block using the guided neural network prediction.
