Binary-neural-network-based lightweight Web AR identification method and system

Abstract

The embodiment of the invention provides a binary-neural-network-based lightweight Web AR identification method and system. Thebinary-neural-network-based lightweight Web AR identification method comprises the steps: loading a target image through a mobile Web browser, preprocessing the image, and sending a target identification task request to an edge server; receiving a binary neural network model and a related executable script returned by the edge server, and performing binary neural network feedforward calculation, obtaining an image identification result, temporarily storing an output result of the sharing layer, judging whether the cross entropy of the image identification result meets a preset threshold; and if not, sending the output result of the sharing layer to the edge serverfor feedforward reasoning. According to the embodiment of the invention, the binary neural network is introduced to accelerate network reasoning, and reduce image recognition loading time delay and equipment energy consumption pressure; computing resources of the mobile terminal are fully utilized, thus effectively relieving computing pressure of the edge server, and a real-time solution for Web AR application is provided.

Description

technical field [0001] The present invention relates to the field of augmented reality technology, and more specifically, to a lightweight Web AR recognition method and system based on a binary neural network. Background technique [0002] Augmented reality (Augmented Reality, referred to as AR) is a technology that seamlessly integrates real world information and virtual world information. It uses computer information to simulate entity information within a certain time and space range, and integrates this virtual information into In the real world, enhance the user's perception of the real world, so as to achieve an effect beyond reality. Augmented reality includes the fusion of various technologies, such as image recognition, 3D modeling, sensor fusion, real-time tracking and registration, and scene fusion. [0003] At present, augmented reality systems mainly use professional equipment (such as head-mounted and glasses equipment) as the core computing equipment to meet ...

AI Technical Summary

Problems solved by technology

However, usually the DNN model is relatively large, with a large amount of calculation, and it is mainly used on the server side, and it is difficult to directly deploy it on the web browser side.

Although some JavaScript inference libraries (such as Tensoflow.js, Keras.js, etc.) implement DNN networks on mobile web browsers, the latency of model loading and network feed-forward inference is still too high, and the energy consumption is high to be effective. Applied to mobile Web AR applications, so exploring a lightweight identification system is one of the important issues to promote WebAR applications

Images