Skip to content

Creating this repository to collect the necessary character image from Bangumi.

Notifications You must be signed in to change notification settings

FacelessManipulator/avatar-gan

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Crawler for Bangumi

Bangumi的Character里面的头像数据很不错,可惜他的爬虫协议不让爬/pic/ 所以分成了两部分,getUrl可以解析图片URL,getImg为了遵守协议,就不实现了。 采用url手动收集了一个二次元头像数据集(混入了一些三次元),因为手有点酸只收集了1000多张,有更多数量需求的请联系Bangumi修改robots.txt
百度云

DCGAN with keras

尝试用Keras实现了一下DCGAN,Loss_D一如既往地远小于Loss_G,正好可以跟下面的WGAN对比。
Keras 2.0.2的同学跑的时候可能会出现下面Error,猜测是Keras旧版本中融合模型的数据传递有损坏,将Keras升级为2.0.4就解决了 InvalidArgumentError (see above for traceback): You must feed a value for placeholder tensor 'dense_4_input' with dtype float [[Node: dense_4_input = Placeholder[dtype=DT_FLOAT, shape=[], _device="/job:localhost/replica:0/task:0/cpu:0"]()]]

DCGAN with pytorch

模型是直接从pytorch-example里面抄过来的,对比Keras那个模型有很多可以借鉴的地方,虽然肉眼上torch这个模型表现更差一点。考虑GAN的历史进程,调參改模型的时间还是用来尝试实现WGAN或者Info之类的扩展更好一点。此模型在cifar10上表现还不错,可惜到了这边就萎了,果然参数跟数据集有很大关系

About

Creating this repository to collect the necessary character image from Bangumi.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published