MTCNN-Tensorflow icon indicating copy to clipboard operation
MTCNN-Tensorflow copied to clipboard

Reproduce MTCNN using Tensorflow

Results 103 MTCNN-Tensorflow issues
Sort by recently updated
recently updated
newest added

image_batch_array, label_batch_array, bbox_batch_array,landmark_batch_array = sess.run([image_batch, label_batch, bbox_batch,landmark_batch]) 卡在了这一步,没有任何报错,或显存不足

您好,感谢开源代码。1 1。代码中关于pnet的损失是: loss = -tf.log(label_prob+1e-10) 跟论文中的cross_entropy = -y*log(p) + (1-y) * (1-log(p)) 好像不一样啊? 当y=1时两者确实相同,但当y = 0时,就不同了 2.pnet 中我训练时,label中好像没有 -2(landmark样本)?

想请问一下wider_face_train_bbx_gt.txt这个文件里面的内容都是什么意思,前面好像是图片的左上角坐标和图片的宽高,但后面的六位数有些看不懂

比如我只训练 P-net,我第一次训练了500张图片,第二次训练了200张图片.我训练第二次的时候Pnet的参数是根据的700张图片还是200张图片?

运行train.py训练ONet,循环617次出现错误tf.errors.OutOfRangeError 测试后发现,错误出现在读取TFRecords文件 image_batch_array, label_batch_array, bbox_batch_array,landmark_batch_array = sess.run([image_batch, label_batch, bbox_batch, landmark_batch]) 上网查了一下,有可能是因为我的正样本太少(pos 1528,part 19753,neg 517384)。 https://blog.csdn.net/pkokocl/article/details/78682324 网上的解决方案提到要修改tfrecord_utils.py文件中的 filename_queue = tf.train.string_input_producer([tfr_filename], shuffle = False, num_epochs = 10) 设置num_epochs,https://www.cnblogs.com/code-wangjun/p/8824406.html 试了一下,没改对。 在train.py中,我注释了以下语句 if coord.should_stop():...

Hello, I have setup my environment as was mentioned in this repository (tensorflow 1.2.1, Cuda 8.0, etc.) However, after the data preparation step, the train_PNet.py code **doesn't train. It just...

我怎么也找不到滑动窗口获取推荐框的代码,请问一下在哪儿?

Hello, I'd like to freeze the graph and save them as pb Currently I've tried to dump the p-net and set "conv4_1/BiasAdd" and "conv4_2/BiasAdd" as output nodes and it can...

怎么将MTCNN_model文件夹 里的三个网络模型文件转换成三个网络对应的.bin和.param文件