广东东软学院本科毕业设计（论文） 基于 Keras 框架下的猫脸特效算法设计 Design of cat face effect algorithm based on Keras 中文摘要随着计算机、数码影像及网络技术的发展, 特别是近年来智能手机的普及和移动通信技术的发展，人们可以简单快捷地获取、存储和传输数码照片， 同时人们对于照片的美观度也有了更高的要求,，尤其是人脸图像照片，越来越多的人开始重视自己的娱乐空间，就像最近很火的抖音短视频和火山小视频等等，其可以通过对小视频的美化，从而增加娱乐的成分，例如人像美化，动物美化等等。另外一点是自从神经网络火起来的时候，越来越多的循环神经网络，卷积神经网络等等模型出现，例如现阶段很火的目标检测算法 Yolo，SSD，R-CNN，图像识别算法MobileNet，ResNet，GoogleNet 等等，本文实现的是一种目标检测算法和图像识别算法结合，实现视频的实时美化。关键字：深度学习 图像识别 目标检测 视频美化广东东软学院本科毕业设计（论文）AbstractWith the development of computer, digital video and network technology, especially in recent years, the popularity of smartphones and the development of mobile communication technology, people can simple fast acquisition, storage, and transmission of digital photos, at the same time for photos of beautiful people also have higher requirements, and especially the face image photos, more and more people begin to pay attention to your own entertainment space, as recently very fire trill short video and volcanic small video, and so on, which can be through to the small video beautification, thereby increasing entertainment elements, such as portrait beautification, animal beautification, and so on.Fire up another point is since the neural network, a growing number of neural network, convolution neural network model and so on, such as a fire at the present stage target detection algorithm Yolo, SSD, R - CNN, MobileNet image recognition algorithm, ResNet, GoogleNet, etc., this paper implemented a target detect...
