karodievas / car-sound-classification-with-keras Goto Github PK

Hi, may I know if how shall I categorize the original source of audio? And where do you get these original source of audio?

I hope I get your help guidance as I am doing this for my Final Year Project and I am a newbie in deep learning.

Thanks in advance.

Hi there,

I'm trying to use my Raspberry Pi to build a sound classifier based on your project. I got some trouble while trying to run the TrainModel.py after Main.py finished

• K.set_image_dim_ordering("th") I searched forum and it's version problem. Replace it with K.image_data_format()=='channels_first' and it works
  --Then I got errors at line 22: Model = kerasModel.get_model(img_width,img_height)
  shows valueError: Negative dimension size caused by subtracting 2 from 1 fro 'max_pooling2d_1/MaxPool' (op:'MaxPool')with input shapes: [?,1,501,32]

Could you help me with that? I used around 200 samples and already change the number of train samples and validation samples

Hi, I have been working on the channel ordering issue and finally I managed to train the model but the results are very bad. Do you have any suggestions on how shall I improve the results?

The results are as attached below:

https://colab.research.google.com/drive/1xHghjvOEVZc3bGtLfgRNw6Xq51d7U_XA

https://colab.research.google.com/drive/1G_oh_f-2TxF8oa6bVpn-qXlRyOfAPAAq

Hi, I have changed the script. The first epoch is successfully run but the same value error appeared.

ValueError: Error when checking input: expected conv2d_1_input to have shape (3, 496, 369) but got array with shape (496, 369, 3)

I have tried running the codes on Jupyter and apparently this is the part where it has errors.

history = model.fit_generator( train_generator, steps_per_epoch=nb_train_samples/batch_size, epochs=nb_epochs, validation_data=validation_generator, validation_steps=nb_validation_samples/batch_size, callbacks=[check_pointer])

Do I simply change the channel orders to channels last instead of channels first? Or is that a better way?

I believe this error is regarding channel ordering but I can't find any way to resolve it after searching on the Internet.

Can you help me with this issue? Which part of the code do I need amendments?

KerasModel.pdf

TrainModel.pdf

Your project seems interesting. Is it actually public ? There's absolutely no documentation on how to run it.

According to the code are these:
img_path = sys.argv[1]
weights = sys.argv[2]
So not just the img path as provided in the readme. Please provide a full example.

Hi,

May I know how shall I properly use the Predictor.py? I tried the following ways but it seems not working.

Hello,

Do you have any idea which loss function shall I use if I have 4 folders for 4 different car brands? Can I use the same binary_crossentropy function or shall I change it to Multi-Class Cross-Entropy?

I appreciate much if you reply.

With regards,
Teoh

$ python predictor.py

Using TensorFlow backend.
Traceback (most recent call last):
File "predictor.py", line 21, in
model.add(MaxPooling2D(pool_size=(2, 2)))
File "/Users/pato/anaconda/lib/python2.7/site-packages/keras/models.py", line 332, in add
output_tensor = layer(self.outputs[0])
File "/Users/pato/anaconda/lib/python2.7/site-packages/keras/engine/topology.py", line 572, in call
self.add_inbound_node(inbound_layers, node_indices, tensor_indices)
File "/Users/pato/anaconda/lib/python2.7/site-packages/keras/engine/topology.py", line 635, in add_inbound_node
Node.create_node(self, inbound_layers, node_indices, tensor_indices)
File "/Users/pato/anaconda/lib/python2.7/site-packages/keras/engine/topology.py", line 166, in create_node
output_tensors = to_list(outbound_layer.call(input_tensors[0], mask=input_masks[0]))
File "/Users/pato/anaconda/lib/python2.7/site-packages/keras/layers/pooling.py", line 160, in call
dim_ordering=self.dim_ordering)
File "/Users/pato/anaconda/lib/python2.7/site-packages/keras/layers/pooling.py", line 210, in _pooling_function
pool_mode='max')
File "/Users/pato/anaconda/lib/python2.7/site-packages/keras/backend/tensorflow_backend.py", line 2866, in pool2d
x = tf.nn.max_pool(x, pool_size, strides, padding=padding)
File "/Users/pato/anaconda/lib/python2.7/site-packages/tensorflow/python/ops/nn_ops.py", line 1793, in max_pool
name=name)
File "/Users/pato/anaconda/lib/python2.7/site-packages/tensorflow/python/ops/gen_nn_ops.py", line 1598, in _max_pool
data_format=data_format, name=name)
File "/Users/pato/anaconda/lib/python2.7/site-packages/tensorflow/python/framework/op_def_library.py", line 763, in apply_op
op_def=op_def)
File "/Users/pato/anaconda/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 2397, in create_op
set_shapes_for_outputs(ret)
File "/Users/pato/anaconda/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 1757, in set_shapes_for_outputs
shapes = shape_func(op)
File "/Users/pato/anaconda/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 1707, in call_with_requiring
return call_cpp_shape_fn(op, require_shape_fn=True)
File "/Users/pato/anaconda/lib/python2.7/site-packages/tensorflow/python/framework/common_shapes.py", line 610, in call_cpp_shape_fn
debug_python_shape_fn, require_shape_fn)
File "/Users/pato/anaconda/lib/python2.7/site-packages/tensorflow/python/framework/common_shapes.py", line 675, in _call_cpp_shape_fn_impl
raise ValueError(err.message)
ValueError: Negative dimension size caused by subtracting 2 from 1 for 'MaxPool' (op: 'MaxPool') with input shapes: [?,1,254,32].