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

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


==Organization Name==

[[:Category:tencent technology (shenzhen) company limited|tencent technology (shenzhen) company limited]]


[[Category:tencent technology (shenzhen) company limited]]

==Inventor(s)==

[[:Category:Zhiyuan Tang of Shenzhen (CN)|Zhiyuan Tang of Shenzhen (CN)]][[Category:Zhiyuan Tang of Shenzhen (CN)]]

[[:Category:Shen Huang of Shenzhen (CN)|Shen Huang of Shenzhen (CN)]][[Category:Shen Huang of Shenzhen (CN)]]

[[:Category:Shidong Shang of Shenzhen (CN)|Shidong Shang of Shenzhen (CN)]][[Category: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.

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

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

### Benefits:
* Improved accuracy in recognition tasks.
* Enhanced feature extraction capabilities.
* Scalability for handling complex media resources.

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

### 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?

### 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.



[[Category:G06N3/08]]
[[Category:G06N3/04]]
[[Category:CPC_G06N3/08]]