tensorflow 能做什么有趣的事情呢?

V2EX = way to explore

V2EX 是一个关于分享和探索的地方

现在注册

已注册用户请登录

这是一个创建于 2783 天前的主题，其中的信息可能已经有所发展或是发生改变。

刚才用这个( https://godoc.org/github.com/tensorflow/tensorflow/tensorflow/go)编译出程序，可以识别一些简单的图片，但是识别美女竟然识别成了尿布，就是这张图 https://timgsa.baidu.com/timg?image&quality=80&size=b9999_10000&sec=1492673172523&di=6da99d8cc334951d8eca387227062c32&imgtype=0&src=http%3A%2F%2Fwww.xbfqw.com%2Fuploadfile%2F2017%2F0309%2F20170309062329245.jpg

TensorFlow

有趣

刚才

事情

5 条回复 • 2017-04-20 14:23:55 +08:00

iot

2017-04-20 12:45:44 +08:00

Desktop ./app -dir=models -image=images/7.jpeg
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations.
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.
BEST MATCH: (99% likely) steam locomotive

Desktop ./app -dir=models -image=images/8.jpeg
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations.
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.
BEST MATCH: (59% likely) electric locomotive

mason961125

2017-04-20 13:15:45 +08:00

1. 训练使用的样本不够多，识别有误很正常。
2. 建议在本地再编译一遍 Tensorflow ，以便能够使用 SSE4.2 和 AVX 指令集，加快运算速度。

iot

2017-04-20 13:17:16 +08:00

@mason961125 还能怎么玩? 能不能做些有用的东西

LeeSeoung

2017-04-20 14:17:44 +08:00

你连你为啥用 tf 都搞不明白，那做些有用的东西，到底对谁有用呢。。。

iot

2017-04-20 14:23:55 +08:00

@LeeSeoung 在知乎大致看了下，不是学这个的，只是玩玩