Jianfeng Chen

Results 19 comments of Jianfeng Chen

It seems that we have encounterd the same problem for real-time face detection. But that part is not my work. I will ask how them solve this problem.

@karloberas This is how we use camera API to achieve real-time detection. You can try it out. ![webwxgetmsgimg](https://user-images.githubusercontent.com/13918606/53022776-d2547580-3496-11e9-8640-e2235fb7451c.jpg)

网络一定会有五个关键点的未知输出的,即使看不到。 你要检测整个脸,可以尝试下修改 [代码里30~32行这3个参数,把阈值调高点](https://github.com/vcvycy/MTCNN4Android/blob/master/app/src/main/java/com/example/vcvyc/mtcnn_new/MTCNN.java#30)

emm... you can use the code for any purpose you want

PB文件你把别人训练好的3个npy文件载入tensorflow,再生成pb就好了。

tf.graph_util.convert_variables_to_constants( sess, input_graph_def, output_node_names, variable_names_whitelist=None, variable_names_blacklist=None ) output_node_names这个参数,在tensorflow 的graph中,不被output_node_names依赖的节点,生成pb文件都会被去掉。所以,你需要在output_node_names把pnet/rnet/onet 所有输出节点都加进去。 @CHANGErIBE

> > > @junwenZhang 参考我的 repo做法 https://github.com/jiangxiluning/facenet_mtcnn_to_mobile > > > > > > 嗨,jiangxiluning ,我想知道的是如何最终合并生成1个pb文件,看起来你那个示例是生成了3个pb文件哈 > > 请问你成功合成1个pb文件了吗 我现在能合成三个pb或者三个npy文件 就是无法合成1个pb文件 合并一次就好了,一个Graph 里面可以有不相连的节点 代码: https://github.com/vcvycy/TensorflowFreezeModel/blob/master/freeze_mtcnn.py