摘要
为了研究不同红外图像量化方法对目标检测网络性能影响的差异,将红外图像量化对深度学习目标检测网络性能影响的研究和分析设计成教学实验.实验内容涉及图像处理、模式识别、计算机视觉等多个专业课程.实验过程包括红外图像量化、网络模型训练、测试分析等多个环节,贯穿基于深度学习的高层视觉任务开发全流程.该实验紧跟学科前沿,促进学生科研能力和综合素质的培养.
Abstract
In order to investigate the impact of different quantization methods on the performance of the deep learning object detection network,a research-oriented teaching experiment is designed,which involves the research and analysis of the impact of infrared image quantization on deep learning object detection performance.The experiment content covers image processing,pattern recognition,computer vision,and so on.The experiment process involves several stages such as infrared image quantization,network model training,and testing analysis,i.e.,it runs through the entire process of developing high-level visual tasks based on deep learning.This experiment keeps experimental teaching up with the subject frontier,promoting the cultivation of students'research ability and comprehensive quality.
基金项目
华中科技大学实验技术研究项目(2023)(2023184018)
维普
万方数据