Tsun-Yi Yang

Any update on this? I get the same error when I define the custom layer with multiple outputs. TypeError: can't multiply sequence by non-int of type 'list'

Thank you very much. It matched now.

On the otherhand, the amount of parameters of DenseNet is small indeed, but the GPU memory will still be consumed by the complex structure instead of the parameters. By using...

OK. Thanks! Despite I wish Torch can also have such property. (QAQ)

啊，三個輸出是因為我之前在嘗試age estimation時有(classification loss, regression loss, center loss)的組合才有這樣的。另外我這個project是按照 https://kexue.fm/archives/4493 做出來的完整範例，要感謝蘇大。 根據你的error你的l2_loss是(?,?,?,?)還蠻奇怪的。四個維度都變成batchsize的感覺，先檢查一下吧。 至於fit_generator我記得要定義輸入的X和Y是多少 類似這樣 ``` hist = model.fit_generator(generator=data_generator_centerloss(X=[x_train, y_train_a_class_value], Y=[y_train_a_class, random_y_train_a], batch_size=batch_size), steps_per_epoch=train_num // batch_size, validation_data=([x_test,y_test_a_class_value], [y_test_a_class, random_y_test_a]), epochs=nb_epochs, verbose=1, callbacks=callbacks) ```...

Also, by downloading the data, it seems like the show.py for the cad won't work directly? ``` >> python show.py --src ../data/cad_vec/0000 --form h5 ``` ------------ This problem is solved...

Hello @CrossEntropy, It's been a long time since I ran this repo. My suggestion is use smaller batch like 32 or 16, and use a lower version of Tensorflow and...

https://github.com/tensorflow/tensorflow/issues/3290 https://github.com/tensorflow/tensorflow/issues/8101 It seems like tf.nn.moments could possibly return nan. You may pick out the nana from the variance and put zero back in. I assume this would solve the...

Hello, sorry I wasn't really maintaining this repo anymore and I am not familiar with ONNX, but luckily I found someone else has done the ONNX thing for me :))))....

Hello, I don't have the experiment against the batchsize, but we do observe small batchsize is much better for the head pose learning. Comparing to general understanding of the batchsize,...