Multi-Branch Cnn For Multi-Scale Age Estimation
IMAGE ANALYSIS AND PROCESSING (ICIAP 2017), PT II(2017)
摘要
Convolutional Neural Networks (CNNs) attracted growing interest in recent years thanks to their high generalization capabilities that are highly recommended especially for applications working in the wild context. However CNNs rely on a huge number of parameters that must be set during training sessions based on very large datasets in order to avoid over-fitting issues. As a consequence the lack in training data is one of the greatest limits for the applicability of deep networks. Another problem is represented by the fixed scale of the filter in the first convolutional layer that limits the analysis performed through the subsequent layers of the network.This paper proposes a way to overcome these problems by the use of a multi-branch convolutional neural network with a reduced deep. In particular its effectiveness for age group classification has been proved demonstrating how, on the one hand, the reduced deep avoids the over-fitting issues, whereas, on the other hand, the multi-branch structure introduces a parallel multi-scale analysis capable to catch multiple size patterns.
更多查看译文
关键词
CNN, Deep learning, Age estimation
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络