search

themightyoarfish / deepVO

:video_camera: Tensorflow implementation of RCCN visual odometry by Wang et al.
50 stars 17 forks source link

Issue in model.py #1

Closed 27Apoorva closed 6 years ago

27Apoorva commented 6 years ago

ValueError: 'conv{index}' is not a valid scope name

themightyoarfish commented 6 years ago

Provide more information and I'll reopen.

27Apoorva commented 6 years ago

Hi, Thank you for your response. I am sorry for not being specific before. I am trying to run your code for KITTI Dataset Grey scale images and poses. Basically, I have raw KITTI data image and poses. I have set up the datamanager.py to read .png images and .txt poses. When I am trying to train, model.py is throwing error. I am running this command: $ python3 main.py -d /home/appu/DeepLearning/Project/Deep-Learning-for-Point-Cloud/Point\ Cloud/2011_09_26/2011_09_26_drive_0001_extract/ -o Adam -l 0.01 -b 10 -e 3 -m 10 -s 11 This is my output:

Number of batches for training: 1 Number of batches for testing: 1 DataManager found 114 images and poses in dataset. Image shape: (512, 1392, 3)

/home/appu/DeepLearning/Project/deepVO/model.py(84)init() -> with tf.variable_scope('inputs'): (Pdb) c w: inputs/imgs:0 /home/appu/DeepLearning/Project/deepVO/model.py(109)init() -> cnn_activation = self.cnn(stacked_image,ksizes,strides,n_channels,reuse=tf.AUTO_REUSE) (Pdb) c Traceback (most recent call last): File "main.py", line 146, in main() File "main.py", line 81, in main use_flownet=args.flownet is not None) File "/home/appu/DeepLearning/Project/deepVO/model.py", line 109, in init cnn_activation = self.cnn(stacked_image,ksizes,strides,n_channels,reuse=tf.AUTO_REUSE) File "/home/appu/DeepLearning/Project/deepVO/model.py", line 230, in cnn with tf.variable_scope(inner_scope_name): File "/usr/local/lib/python3.4/dist-packages/tensorflow/python/ops/variable_scope.py", line 1903, in enter return self._enter_scope_uncached() File "/usr/local/lib/python3.4/dist-packages/tensorflow/python/ops/variable_scope.py", line 1947, in _enter_scope_uncached current_name_scope_name = current_name_scope.enter() File "/usr/local/lib/python3.4/dist-packages/tensorflow/python/framework/ops.py", line 5775, in enter return self._name_scope.enter() File "/usr/lib/python3.4/contextlib.py", line 59, in enter return next(self.gen) File "/usr/local/lib/python3.4/dist-packages/tensorflow/python/framework/ops.py", line 3979, in name_scope raise ValueError("'%s' is not a valid scope name" % name) ValueError: 'conv{index}' is not a valid scope name

Please tell me what I am doing wrong and if you need any more information. Thank you.

themightyoarfish commented 6 years ago

I see. Python 3.6 is required, since format strings are not supported in previous versions.

On 13 April 2018 20:26:50 CEST, 27Apoorva notifications@github.com wrote:

Hi, Thank you for your response. I am sorry for not being specific before. I am trying to run your code for KITTI Dataset Grey scale images and poses. Basically, I have raw KITTI data image and poses. I have set up the datamanager.py to read .png images and .txt poses. When I am trying to train, model.py is throwing error. I am running this command: $ python3 main.py -d /home/appu/DeepLearning/Project/Deep-Learning-for-Point-Cloud/Point\ Cloud/2011_09_26/2011_09_26_drive_0001_extract/ -o Adam -l 0.01 -b 10 -e 3 -m 10 -s 11 This is my output:

Number of batches for training: 1 Number of batches for testing: 1 DataManager found 114 images and poses in dataset. Image shape: (512, 1392, 3)

/home/appu/DeepLearning/Project/deepVO/model.py(84)init() -> with tf.variable_scope('inputs'): (Pdb) c w: inputs/imgs:0 /home/appu/DeepLearning/Project/deepVO/model.py(109)init() -> cnn_activation = self.cnn(stacked_image,ksizes,strides,n_channels,reuse=tf.AUTO_REUSE) (Pdb) c Traceback (most recent call last): File "main.py", line 146, in main() File "main.py", line 81, in main use_flownet=args.flownet is not None) File "/home/appu/DeepLearning/Project/deepVO/model.py", line 109, in init cnn_activation = self.cnn(stacked_image,ksizes,strides,n_channels,reuse=tf.AUTO_REUSE) File "/home/appu/DeepLearning/Project/deepVO/model.py", line 230, in cnn with tf.variable_scope(inner_scope_name): File "/usr/local/lib/python3.4/dist-packages/tensorflow/python/ops/variable_scope.py", line 1903, in enter return self._enter_scope_uncached() File "/usr/local/lib/python3.4/dist-packages/tensorflow/python/ops/variable_scope.py", line 1947, in _enter_scope_uncached current_name_scope_name = current_name_scope.enter() File "/usr/local/lib/python3.4/dist-packages/tensorflow/python/framework/ops.py", line 5775, in enter return self._name_scope.enter() File "/usr/lib/python3.4/contextlib.py", line 59, in enter return next(self.gen) File "/usr/local/lib/python3.4/dist-packages/tensorflow/python/framework/ops.py", line 3979, in name_scope raise ValueError("'%s' is not a valid scope name" % name) ValueError: 'conv{index}' is not a valid scope name

Please tell me what I am doing wrong and if you need any more information. Thank you.

-- You are receiving this because you modified the open/close state. Reply to this email directly or view it on GitHub: https://github.com/themightyoarfish/deepVO/issues/1#issuecomment-381222442

themightyoarfish commented 6 years ago

Also, be advised we didn't actually run this on the KITTI dataset since we could not even fit our own data into GPU memory with reasonable parameters. I'm not sure if the implementation is correct.