Neural network construction method and device and image processing method and device

Abstract

The invention discloses a neural network construction method and device and an image processing method and device in the field of computer vision in the artificial intelligence field. The neural network construction method comprises the following steps: determining a search space and a plurality of construction units; stacking the plurality of construction units to obtain a search network, with the search network being a neural network for searching a neural network structure; and optimizing a network structure of a construction unit in the search network in the search space, wherein in the optimization process, the search space is gradually reduced, the number of the construction units is gradually increased, and due to the reduction of the search space and the increase of the number of the construction units, the video memory consumption generated in the optimization process is within a preset range; and establishing the target neural network according to the optimized construction unit. Under the condition that video memory resources are certain, the neural network meeting application requirements well can be constructed.

Description

technical field [0001] The present application relates to the field of artificial intelligence, and more specifically, to a neural network construction method, image processing method and device. Background technique [0002] Artificial intelligence (AI) is a theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is the branch of computer science that attempts to understand the nature of intelligence and produce a new class of intelligent machines that respond in ways similar to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines, so that the machines have the functions of perception, reasoning and decision-making. Research in the field of artific...

AI Technical Summary

Problems solved by technology

In the optimization process of this search method, all possible operations will be put into the search space, which leads to the huge memory space required in the optimization process, so it can only be stacked into a shallow search network

However, the final target neural network to be constructed is often deep, which leads to a large depth difference between the search network and the target neural network, and the construction unit obtained by optimizing the shallow search network is not completely suitable for the deep network. The target neural network, so that the final target neural network may not meet the application requirements well

Images