deepsound-project / genre-recognition Goto Github PK

Getting this error during runtime..I have upgraded tensorflow to latest version (1.70). How do i get rid of this error?

how to build au files?

Using TensorFlow backend.
Traceback (most recent call last):
File "model_to_tfjs.py", line 31, in
main(options.model_path, options.output_path)
File "model_to_tfjs.py", line 13, in main
model = load_model(model_path)
File "/home/sprio/anaconda3/lib/python3.6/site-packages/tensorflow/python/keras/engine/saving.py", line 251, in load_model
training_config['weighted_metrics'])
KeyError: 'weighted_metrics'

I did this experiment with your code, but the validation accuracy is only 0.74, how can I improve the val_acc ?

Hello everyone!
I'm new to neural networks and every technology related to it. But I try experimenting and learning new things. I came across this git while I was surfing and thought why not give it a try! So, I've done everything what the author suggested. Right from installing the required packages to running of the server. But when I run "server.py" file, I get the following error:

(C:\Users\lenovo\Anaconda3\envs\new_env) C:\genre-recognition-master>python server.py XXX lineno: 56, opcode: 0 Traceback (most recent call last): File "server.py", line 96, in <module> options.weights_path) File "C:\genre-recognition-master\genre_recognizer.py", line 11, in __init__ model = model_from_yaml(f.read()) File "C:\Users\lenovo\Anaconda3\envs\new_env\lib\site-packages\tensorflow\python\keras\engine\saving.py", line 329, in model_from_yaml return deserialize(config, custom_objects=custom_objects) File "C:\Users\lenovo\Anaconda3\envs\new_env\lib\site-packages\tensorflow\python\keras\layers\serialization.py", line 64, in deserialize printable_module_name='layer') File "C:\Users\lenovo\Anaconda3\envs\new_env\lib\site-packages\tensorflow\python\keras\utils\generic_utils.py", line 173, in deserialize_keras_object list(custom_objects.items()))) File "C:\Users\lenovo\Anaconda3\envs\new_env\lib\site-packages\tensorflow\python\keras\engine\network.py", line 1162, in from_config process_node(layer, node_data) File "C:\Users\lenovo\Anaconda3\envs\new_env\lib\site-packages\tensorflow\python\keras\engine\network.py", line 1120, in process_node layer(input_tensors[0], **kwargs) File "C:\Users\lenovo\Anaconda3\envs\new_env\lib\site-packages\tensorflow\python\keras\engine\base_layer.py", line 703, in __call__ outputs = self.call(inputs, *args, **kwargs) File "C:\Users\lenovo\Anaconda3\envs\new_env\lib\site-packages\tensorflow\python\keras\layers\core.py", line 709, in call return self.function(inputs, **arguments) File "train_model.py", line 56, in <lambda> function=lambda x: K.mean(x, axis=1), SystemError: unknown opcode

I hope someone to address this issue soon and Thanks in advance :)!

I found that with audios sampled at 22050Hz, a FFT window of 2048 is 0.1 seconds, which means 100ms not 10ms as you say in http://deepsound.io/music_genre_recognition.html.

Thanks a lot for this repo and the old documentation.

This code is written in python 2 but tensorflow is not supported in python 2. So what should I do?

Clicking http://deepsound.io/genres/ link on desktop redirects me to a page about downloading a sketchy browser extension. On mobile, I get a popup about winning a contest. Has the site been hacked?

Dear Friend.

This is very nice effort. I would like to ask that how is it possible to run the trained model like demo offline without this link. Please if you can share recipe.
I mean without building like docker. as I cannot see the classifier code how it works. !! Please if you can guide me.

Best regards

Hi，could you give the detailed version information of python ,numpy and scipy ?

Readme talks about a blog post Convolutional-Recurrent Neural Network for Live Music Genre Recognition
http://ww1.deepsound.io/music_genre_recognition.html

But it seems to be broken.

The

code doesn't work until you fix the code on line 16 and 37.
The file extension should be .wav or the respective file extension of your music files.

14 def get_default_shape(dataset_path):
15 tmp_features, _ = load_track(os.path.join(dataset_path,
16 'blues/blues.00000.au'))

35 for (genre_index, genre_name) in enumerate(GENRES):
36 for i in range(TRACK_COUNT // len(GENRES)):
37 file_name = '{}/{}.000{}.au'.format(genre_name,
38 genre_name, str(i).zfill(2))

Which version of Tensor flow does the project use?