Tencent technology (shenzhen) company limited (20240249144). USING SHARED AND NON-SHARED PARAMETERS IN AN ATTENTION MODULE-BASED RECOGNITION MODEL simplified abstract

From WikiPatents
Jump to navigation Jump to search

USING SHARED AND NON-SHARED PARAMETERS IN AN ATTENTION MODULE-BASED RECOGNITION MODEL

Organization Name

tencent technology (shenzhen) company limited

Inventor(s)

Zhiyuan Tang of Shenzhen (CN)

Shen Huang of Shenzhen (CN)

Shidong Shang of Shenzhen (CN)

USING SHARED AND NON-SHARED PARAMETERS IN AN ATTENTION MODULE-BASED RECOGNITION MODEL - A simplified explanation of the abstract

This abstract first appeared for US patent application 20240249144 titled 'USING SHARED AND NON-SHARED PARAMETERS IN AN ATTENTION MODULE-BASED RECOGNITION MODEL

The patent application describes a recognition method that involves inputting a media resource feature into a recognition model with multiple layers of attention modules to obtain a representation vector.

  • The method utilizes layers of attention modules to process the media resource feature and generate a representation vector.
  • Each layer of attention module determines weight parameters and input representation vectors based on shared and non-shared parameters.
  • The representation vector outputted by each layer of attention module is used to determine parameters for the next layer in the model.
  • Shared parameters are utilized by at least two layers of attention modules among the total layers in the model.
      1. Potential Applications:

This technology can be applied in image recognition, speech recognition, video analysis, and other fields requiring pattern recognition and feature extraction.

      1. Problems Solved:

The method addresses the challenge of efficiently processing media resource features to extract meaningful representations for recognition tasks.

      1. Benefits:
  • Improved accuracy in recognition tasks.
  • Enhanced feature extraction capabilities.
  • Scalability for handling complex media resources.
      1. Commercial Applications:

The technology can be utilized in security systems, content recommendation engines, autonomous vehicles, and medical imaging analysis software.

      1. Questions about Recognition Method:

1. How does the use of multiple layers of attention modules improve the recognition process? 2. What are the potential limitations of this method in real-world applications?

      1. Frequently Updated Research:

Stay updated on advancements in attention-based recognition models and their applications in various industries.


Original Abstract Submitted

a recognition method includes inputting a media resource feature into a recognition model having n layers of attention modules, and processing the media resource feature by using the n layers of attention modules to obtain a representation vector. an ilayer of attention module is configured to determine an ilayer weight parameter and an iinput representation vector based on (i) a group of shared parameters and (ii) an igroup of non-shared parameters, and determine an ilayer of representation vector outputted by the ilayer of attention module. when i is less than n, the ilayer of representation vector determines an (i+1)group of non-shared parameters used by an (i+1)layer of attention module and at least two layers of attention modules among the n layers of attention modules share the group of shared parameters.

Retrieved from "http://wikipatents.org/index.php?title=Tencent_technology_(shenzhen)_company_limited_(20240249144)._USING_SHARED_AND_NON-SHARED_PARAMETERS_IN_AN_ATTENTION_MODULE-BASED_RECOGNITION_MODEL_simplified_abstract&oldid=201772"