摘要
随着人工智能和计算机视觉技术的快速发展,图像识别在许多领域中变得尤为重要,而中餐食物的复杂性和多样性给图像分类带来了巨大挑战。ResNet因其高效信息传递能力,在多个领域有广泛应用。现有的ResNet结构在处理中餐食物图像时,可能无法充分应对其独有的特征和复杂性。通过在统一条件下对ResNet结构进行研究,训练和比较不同的ResNet结构,选出最适合中餐食物图像识别的模型,并进一步优化网络性能。实验结果表明,本实验中探索到的ResNet结构在提高分类准确率方面具有显著优势。研究结果为中餐图像识别领域的应用提供了新的技术路径和理论支持。
With the rapid development of Artificial Intelligence and Computer Vision technology,image recognition has become particularly important in many fields,and the complexity and diversity of Chinese food have brought great challenges to image classification.ResNet is widely used in many fields due to its efficient information transmission capability.The existing ResNet structure may not adequately cope with the unique characteristics and complexity of Chinese food images.By studying the ResNet structure under the unified conditions,training and comparing different ResNet structures,the most suitable model for Chinese food image recognition is selected,and the network performance is further optimized.The experimental results show that the ResNet structure explored in this experiment has significant advantages in improving the classification accuracy.The research results provide a new technical path and theoretical support for the application of Chinese food image recognition.
作者
施晨炜
SHI Chenwei(School of Information Science and Engineering,East China University of Science and Technology,Shanghai 200237,China)
出处
《现代信息科技》
2025年第3期73-78,共6页
Modern Information Technology
基金
上海市大学生创新创业训练计划项目(S202310251138)。
关键词
图像识别
残差神经网络
中餐
神经网络
人工智能
image recognition
ResNet
Chinese food
Neural Network
Artificial Intelligence
