查询结果:   顾郑平,朱敏.基于深度学习的鱼类分类算法研究[J].计算机应用与软件,2018,35(1):200 - 205.
中文标题
基于深度学习的鱼类分类算法研究
发表栏目
人工智能与识别
摘要点击数
1198
英文标题
FISH CLASSIFICATION ALGORITHM BASED ON DEPTH LEARNING
作 者
顾郑平 朱敏 Gu Zhengping Zhu Min
作者单位
华东师范大学计算机科学与软件工程学院计算中心 上海 200062     
英文单位
Computer Center,School of Computer Science and Software Engineering,East China Normal University,Shanghai 200062,China     
关键词
深度学习 卷积神经网络 迁移学习 支持向量机
Keywords
Deep learning Convolutional neural network Transfer learning SVM
基金项目
作者资料
顾郑平,硕士生,主研领域:智能计算与智能系统。朱敏,教授级高工。 。
文章摘要
回顾近年来国内外对鱼类分类的研究进展,指出传统方法存在的缺陷。深度学习是目前图像分类的主流方法。研究基于卷积神经网络CNN(Convolutional Neural Network)的鱼类分类模型,并以该模型为基础,进一步提出利用迁移学习,以预训练网络的特征结合SVM算法(PreCNN+SVM)的混合分类模型。实验以Fish4-Knowledge (F4K) 作为数据集,使用TensorFlow训练网络模型。实验结果表明,利用PreCNN+SVM算法,取得了98.6%的准确率,较传统方法有显著提高。对于小规模数据集,有效解决了需要人工提取特征的不可迁移性。
Abstract
Reviewing the research progress of fish classification at home and abroad in recent years, the shortcomings of the traditional methods are pointed out. Deep learning is the mainstream method of image classification. This paper studied the fish classification model based on Convolutional Neural Network (CNN). Based on this model, we proposed a hybrid model based on transfer learning algorithm using pre-training neural network and SVM algorithm (PreCNN+SVM). The experimental results showed that using PreCNN + SVM algorithm with Fish4-Knowledge (F4K) data sets and TensorFlow model, the accuracy of 98.6% was achieved, which was significantly higher than the traditional method. For small-scale data sets, it effectively solved the immutability that need to extract features manually.
下载PDF全文