用深度学习做的机械抓取的目标识别，对桌面上的问题进行检测并识别出抓取的区域以及抓取方式-Learning to do with the depth of the object recognition mechanical crawl on the issues on the desktop to detect and identify the areas and crawl crawl way

一种专门为移动端设备而设计的高效卷积神经网络结构——ShuffleNet，这种新结构将点态组卷积（pointwise group convolution）和通道随机混合（channel shuffle）这两种经典方法进行结合与改进，大大提升了计算效率 。(ShuffleNet is an efficient convolutional neural network designed for mobile terminal devices

基于深度学习的目标检测方法，效果很好，可以进行学习(Object detection method based on deep learning)