17430644. METHOD AND SYSTEM OF DNN MODULARIZATION FOR OPTIMAL LOADING simplified abstract (SAMSUNG ELECTRONICS CO., LTD.)

From WikiPatents
Jump to navigation Jump to search

METHOD AND SYSTEM OF DNN MODULARIZATION FOR OPTIMAL LOADING

Organization Name

SAMSUNG ELECTRONICS CO., LTD.

Inventor(s)

Brijraj Singh of Bengaluru (IN)

Mayukh Das of Bengaluru (IN)

Yash Hemant Jain of Bengaluru (IN)

Sharan Kumar Allur of Bengaluru (IN)

Venkappa Mala of Bengaluru (IN)

Praveen Doreswamy Naidu of Bengaluru (IN)

METHOD AND SYSTEM OF DNN MODULARIZATION FOR OPTIMAL LOADING - A simplified explanation of the abstract

This abstract first appeared for US patent application 17430644 titled 'METHOD AND SYSTEM OF DNN MODULARIZATION FOR OPTIMAL LOADING

Simplified Explanation

The abstract describes a method for optimizing the loading of a deep neural network (DNN) model on an electronic device. The method involves receiving the DNN model, obtaining parameters associated with the device and the model, determining the number of sub-models and a splitting index based on these parameters, and splitting the DNN model into multiple sub-models accordingly.

  • The method aims to modularize a DNN model for efficient execution on an electronic device.
  • The electronic device receives a DNN model and parameters associated with both the device and the model.
  • Based on these parameters, the method determines the number of sub-models and a splitting index.
  • The DNN model is then split into multiple sub-models according to the determined parameters.

Potential Applications

  • This method can be applied in various fields where DNN models are used, such as computer vision, natural language processing, and speech recognition.
  • It can be beneficial in resource-constrained devices like smartphones, IoT devices, and embedded systems, where optimizing the loading of DNN models is crucial.

Problems Solved

  • Loading and executing large DNN models on electronic devices with limited resources can be challenging.
  • This method solves the problem of efficiently utilizing the available resources by modularizing the DNN model and optimizing its loading.

Benefits

  • By splitting the DNN model into sub-models, the method allows for better resource management and utilization.
  • It enables the execution of complex DNN models on devices with limited computational power and memory.
  • The modularization approach improves the overall performance and efficiency of DNN model execution on electronic devices.


Original Abstract Submitted

A method of deep neural network (DNN) modularization for optimal loading includes receiving, by an electronic device, a DNN model for execution, obtaining, by the electronic device, a plurality of parameters associated with the electronic device and a plurality of parameters associated with the DNN model, determining, by the electronic device, a number of sub-models of the DNN model and a splitting index, based on the obtained plurality of parameters associated with the electronic device and the obtained plurality of parameters associated with the DNN model, and splitting, by the electronic device, the received DNN model into a plurality of sub-models, based on the determined number of sub-models of the DNN model and the determined splitting index.

Retrieved from "http://wikipatents.org/index.php?title=17430644._METHOD_AND_SYSTEM_OF_DNN_MODULARIZATION_FOR_OPTIMAL_LOADING_simplified_abstract_(SAMSUNG_ELECTRONICS_CO.,_LTD.)&oldid=22960"