18479611. INDEPENDENT POSITIONING OF AUXILIARY INFORMATION IN NEURAL NETWORK BASED PICTURE PROCESSING simplified abstract (HUAWEI TECHNOLOGIES CO., LTD.)

From WikiPatents
Jump to navigation Jump to search

INDEPENDENT POSITIONING OF AUXILIARY INFORMATION IN NEURAL NETWORK BASED PICTURE PROCESSING

Organization Name

HUAWEI TECHNOLOGIES CO., LTD.

Inventor(s)

Timofey Mikhailovich Solovyev of Munich (DE)

Elena Alexandrovna Alshina of Munich (DE)

Biao Wang of Shenzhen (CN)

Alexander Alexandrovich Karabutov of Munich (DE)

Mikhail Vyacheslavovich Sosulnikov of Munich (DE)

Georgy Petrovich Gaikov of Munich (DE)

Han Gao of Shenzhen (CN)

Panqi Jia of Munich (DE)

Esin Koyuncu of Munich (DE)

Sergey Yurievich Ikonin of Moscow (RU)

Semih Esenlik of Munich (DE)

INDEPENDENT POSITIONING OF AUXILIARY INFORMATION IN NEURAL NETWORK BASED PICTURE PROCESSING - A simplified explanation of the abstract

This abstract first appeared for US patent application 18479611 titled 'INDEPENDENT POSITIONING OF AUXILIARY INFORMATION IN NEURAL NETWORK BASED PICTURE PROCESSING

Simplified Explanation

The patent application describes methods and apparatuses for processing picture data using a neural network with two or more layers, particularly in the field of neural network-based video compression technologies. The processing involves combining two kinds of data obtained from different stages of processing by the network, leading to greater scalability and a more flexible design of the neural network architecture for improved encoding/decoding performance.

  • Neural network-based processing of picture data
  • Utilizes two or more layers in the network
  • Combines two kinds of data from different processing stages
  • Improves scalability and flexibility of network design
  • Enhances encoding/decoding performance

Potential Applications

This technology can be applied in various fields such as video compression, image processing, artificial intelligence, and data analysis.

Problems Solved

1. Improved scalability and flexibility in neural network architecture 2. Enhanced encoding/decoding performance in picture data processing

Benefits

1. Greater efficiency in processing picture data 2. Enhanced performance in video compression technologies 3. Improved design flexibility in neural networks

Potential Commercial Applications

1. Video streaming services 2. Image editing software 3. Surveillance systems

Possible Prior Art

Prior art in the field of neural network-based video compression technologies may include research papers, patents, and existing products that utilize similar techniques for processing picture data.

What are the specific neural network architectures used in this technology?

The specific neural network architectures used in this technology are not detailed in the abstract. Further information on the types of neural networks employed and their configurations would provide a clearer understanding of the processing methods.

How does the combination of two kinds of data from different processing stages improve encoding/decoding performance?

The abstract mentions that combining two kinds of data from different processing stages leads to greater scalability and a more flexible design of the neural network architecture. However, the specific mechanisms through which this combination enhances encoding/decoding performance are not explicitly explained. Further details on the relationship between the combined data and performance improvements would provide a deeper insight into the technology.


Original Abstract Submitted

This application provides methods and apparatuses for processing of picture data or picture feature data using a neural network with two or more layers. The present disclosure may be applied in the field of artificial intelligence (AI)-based video or picture compression technologies, and in particular, to the field of neural network-based video compression technologies. According to some embodiments, two kinds of data are combined during the processing including processing by the neural network. The two kinds of data are obtained from different stages of processing by the network. Some of the advantages may include greater scalability and a more flexible design of the neural network architecture which may further lead to better encoding/decoding performance.