Year: 2024
Type: article
Abstract: This paper studies the optimization method of deep learning algorithm in real-time image processing, focusing on the optimization strategies of model compression, hardware acceleration and data processing. Through these technologies, the reasoning speed of the image processing system is significantly improved, while the high accuracy of the model is maintained. The research shows that these optimization strategies have wide application potential in resource-constrained environments, providing efficient solutions for practical applications.
Authors: Genjuan Ma, Yan Li
Institution: Communication University of China, Nanjing
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4405072644