alpr-unconstrained
alpr-unconstrained copied to clipboard
sergiomsilva
Create custom model for 230.000 images
I have jpg files that contains 230.000 car plate images for graduate project and i'm new for this topic. What exactly should I do for plate recognation. Which format should I use and how can I convert 230.000 images to data format that you will say.
Share the knowledge here ! :D i'm listening also, watching forks ...
@Frank-qlu do you have annotations
i have some annotations,but my tranning have some problem
What kind of problem
python2 train-detector.py --model models/eccv-model-scracth --name my-trained-model --train-dir samples/train-detector --output-dir models/my-trained-model/ -op Adam -lr .001 -its 30 -bs 64
Using TensorFlow backend.
2019-11-26 11:32:38.850268: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 2799995000 Hz
2019-11-26 11:32:38.854033: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x5c7ece0 executing computations on platform Host. Devices:
2019-11-26 11:32:38.854058: I tensorflow/compiler/xla/service/service.cc:175] StreamExecutor device (0): Host, Default Version
Traceback (most recent call last):
File "train-detector.py", line 74, in
Share the knowledge here ! :D i'm listening also, watching forks ...
Using TensorFlow backend. 2019-11-18 14:10:11.658100: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA 2019-11-18 14:10:11.679120: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 3192000000 Hz 2019-11-18 14:10:11.680140: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x563b23931330 executing computations on platform Host. Devices: 2019-11-18 14:10:11.680166: I tensorflow/compiler/xla/service/service.cc:175] StreamExecutor device (0): Host, Default Version Traceback (most recent call last): File "train-detector.py", line 74, in model,model_stride,xshape,yshape = load_network(args.model,dim) File "train-detector.py", line 31, in load_network output_shape = tuple([s.value for s in outputs.shape[1:]]) AttributeError: 'int' object has no attribute 'value' Could please you tell why I'm getting this error and how to fix it?
Share the knowledge here ! :D i'm listening also, watching forks ...
Using TensorFlow backend. 2019-11-18 14:10:11.658100: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA 2019-11-18 14:10:11.679120: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 3192000000 Hz 2019-11-18 14:10:11.680140: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x563b23931330 executing computations on platform Host. Devices: 2019-11-18 14:10:11.680166: I tensorflow/compiler/xla/service/service.cc:175] StreamExecutor device (0): Host, Default Version Traceback (most recent call last): File "train-detector.py", line 74, in model,model_stride,xshape,yshape = load_network(args.model,dim) File "train-detector.py", line 31, in load_network output_shape = tuple([s.value for s in outputs.shape[1:]]) AttributeError: 'int' object has no attribute 'value' Could please you tell why I'm getting this error and how to fix it?
did you solve this problem ? @sergiomsilva
how can we solve the error i get the same error
Had the same problem with python 3.7
I didn't fully test it yet (takes to long without GPU), but at least the error above is gone and the Iterations are running now.
Made the following changes:
train-detector.py
Line 31:
output_shape = tuple([s for s in outputs.shape[1:]])
Line 102:
Ytrain = np.empty((batch_size, int(dim/model_stride), int(dim/model_stride), 2*4+1))
src/sampler.py
Line 15:
outsize = int(dim/stride)
src/loss.py
Line 8:
Pred = -tf.math.log(Pred)
