公开(公告)号：US20230068381A1

公开(公告)日：2023-03-02

申请号：US17961453

申请日：2022-10-06

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Tejpratap Venkata Subbu Lakshmi GOLLANAPALLI ,  Arun ABRAHAM ,  Raja KUMAR ,  Pradeep NELAHONNE SHIVAMURTHAPPA ,  Vikram Nelvoy RAJENDIRAN ,  Prasen Kumar SHARMA

IPC: G06N3/08 ,  G06V10/82

Abstract: Various embodiments of the disclosure disclose a method for quantizing a Deep Neural Network (DNN) model in an electronic device. The method includes: estimating, by the electronic device, an activation range of each layer of the DNN model using self-generated data (e.g. retro image, audio, video, etc.) and/or a sensitive index of each layer of the DNN model; quantizing, by the electronic device, the DNN model based on the activation range and/or the sensitive index; and allocating, by the electronic device, a dynamic bit precision for each channel of each layer of the DNN model to quantize the DNN model.