i cannot find the result picture

[Lvhhhh]

thanks for your work! i download it and want to test it performance. but after the command. i found no picture on my result folder. this is my command, "python test.py -data ./pic" (the 'pic ' is the folder where i put my test pictures in 'left' and 'right',such as"./pic/left/0l.png"and "./pic/right/0r.png")

the content on the 'test_params.json' are: "{ "pspath": "./res", "batch_size": 1, "w_path": "./model/pretrained_model_weight.hdf5", "max_q_size": 3, "verbose": 1 } " the content on the 'util_param.json' are: "{ "crop_width": 128, "crop_height": 96, "new_max": 1, "new_min": -1, "old_max": 256, "old_min": 0, "val_ratio": 0.1, "file_extension": "png", "seed": 1234, "fraction": 1 } " thank you for your help!

[LinHungShi]

Hi, could you please post parts of the message?
Btw, I am still training the model with the SceneFlow dataset, so the result might not be good.

[Lvhhhh]

now. it have these problem sing TensorFlow backend. 2017-09-16 10:22:56.473879: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations. 2017-09-16 10:22:56.473910: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations. 2017-09-16 10:22:56.473918: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations. 2017-09-16 10:22:56.601133: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero 2017-09-16 10:22:56.601492: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties: name: GeForce GTX 1060 6GB major: 6 minor: 1 memoryClockRate (GHz) 1.7845 pciBusID 0000:01:00.0 Total memory: 5.93GiB Free memory: 5.58GiB 2017-09-16 10:22:56.601512: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 2017-09-16 10:22:56.601521: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y 2017-09-16 10:22:56.601532: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GeForce GTX 1060 6GB, pci bus id: 0000:01:00.0) Predict data using generator... Exception in thread Thread-1: Traceback (most recent call last): File "/usr/lib/python2.7/threading.py", line 801, in __bootstrap_inner self.run() File "/usr/lib/python2.7/threading.py", line 754, in run self.__target(*self.__args, **self.__kwargs) File "/usr/local/lib/python2.7/dist-packages/keras/utils/data_utils.py", line 568, in data_generator_task generator_output = next(self._generator) StopIteration

Traceback (most recent call last): File "test.py", line 49, in pred = model.predict_generator(generator, max_queue_size = max_q_size, steps = bs, verbose = verbose) File "/usr/local/lib/python2.7/dist-packages/keras/legacy/interfaces.py", line 87, in wrapper return func(*args, **kwargs) File "/usr/local/lib/python2.7/dist-packages/keras/engine/training.py", line 2272, in predict_generator generator_output = next(output_generator) StopIteration

[LinHungShi]

Hi, please set max_q_size to 1, then test your model again. Show me the error message if it still occurs.

[Lvhhhh]

the test_params.json { "data":"./pic", "pspath": "./prediction", "batch_size": 1, "w_path": "model_weight.hdf5", "max_q_size": 1, "verbose": 1 }

the error message:

Using TensorFlow backend. Loading pretrained cost weight... 2017-09-18 13:54:29.217078: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations. 2017-09-18 13:54:29.217100: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations. 2017-09-18 13:54:29.217107: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations. 2017-09-18 13:54:29.342172: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero 2017-09-18 13:54:29.342498: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties: name: GeForce GTX 1060 6GB major: 6 minor: 1 memoryClockRate (GHz) 1.7845 pciBusID 0000:01:00.0 Total memory: 5.93GiB Free memory: 5.59GiB 2017-09-18 13:54:29.342513: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 2017-09-18 13:54:29.342519: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y 2017-09-18 13:54:29.342528: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GeForce GTX 1060 6GB, pci bus id: 0000:01:00.0) Predict data using generator... Exception in thread Thread-1: Traceback (most recent call last): File "/usr/lib/python2.7/threading.py", line 801, in __bootstrap_inner self.run() File "/usr/lib/python2.7/threading.py", line 754, in run self.__target(*self.__args, **self.__kwargs) File "/usr/local/lib/python2.7/dist-packages/keras/utils/data_utils.py", line 568, in data_generator_task generator_output = next(self._generator) StopIteration

1/1 [==============================] - 1s

Process finished with exit code 0

[Lvhhhh]

1.i use the newest version of your code. when i python train.py i found the loss is bigger and bigger. train_params.json are followed: { "weight_save_path": "model/model_weight.hdf5", "period": 1, "verbose": 1, "log_save_path": "log", "max_q_size": 1, "save_best_only": 1, "weight_path": "model/pretrained_model_weight.hdf5", "learning_rate": 0.001, "batch_size": 1, "epochs": 1, "epsilon": 0.00000001, "rho": 0.9, "decay": 0.0, "loss_function": "mean_absolute_error", "cost_volume_weight_save_path": "model/cost_weight.hdf5", "cost_volume_weight_path": "model/cost_weight.hdf5", "linear_output_weight_save_path": "model/linear_output_weight.hdf5", "linear_output_weight_path": "model/linear_output_weight.hdf5",
"pspath": "./prediction", "psdata":"./pic" }

here are the environment.json { "sceneflow_root": "/home/lvhao/data/", "driving_root": "driving", "driving_train": "frames_cleanpass", "driving_label": "disparity", "monkaa_root": "monkaa", "monkaa_train": "frames_cleanpass", "monkaa_label": "disparity", "train_all": 0, "train_driving": 1, "train_monkaa": 0 } (i just use the driving sample) 2.in test.py i make some changes: add a line"psdata = tp['psdata']" behind the line23"pspath = tp['pspath']" and change the line 27 to "parser.add_argument('-data', help = 'data used for prediction', default = psdata)" and i set the left and right test pictures such as "./pic/left/0_left.png"and "./pic/right/0_right.png". (because i want to just input the 'python test.py') here are all changes i make . i dont know why it does not work! thank you for your time!

[Lvhhhh]

the message of util_params.json are here: { "crop_width": 128, "crop_height": 128, "val_ratio": 0.1, "file_extension": "png", "seed": 1234, "fraction": 1 }

[LinHungShi]

1. Hi, the error occurs because you feed only one data sample. However, you should see the prediction result in the directory.
2. What value do you set for pspath?
3. Since the model is trained with patches randomly cropped from images, it requires lots of epochs before the convergence. I am still training the model with the Driving and Monkaa dataset, and will upload the result to Github as soon as possible.

[Lvhhhh]

hi. after i copy the '0l.png' to '0l.png' and '1l.png'(the same as the right images) .the error message disappered . i can use PIL to show the result image~ now i want to know:why only one data sample make error message? thanks! and i am looking forward to your training result!!

[guandog]

six six six