dohoseok/context-based-parking-slot-detect

context-based-parking-slot-detect's Issues

What's the label tool

Hello， Author：
I’d like to ask about the labeling tool. How can I label the angle from the label txt?
Thanks.

How to prepare the dataset ps2.0 and train the model with it

Hello, Author:
I want to train the dataset ps2.0 and compare the model with others. But I don't know how to prepare the dataset, could you do me a favor? Thanks very much.

how to generate labels of ps2.0

Hi author, may I ask how to generate the labels of ps2.0, since the original labels of ps2.0 don't fit the input format

Regarding yolov3.ckpt file

Hello, during implementation I got an error for yolov3.ckpt file.
That file is not present anywhere. Can you please help me with that

I want to ask what is your version of tensorflow

Machine stuck?

I am at step 3 in "Prepare Dataset". And after converting the images it has just this output:

Finished converting the parking_space dataset:
file_idx = 0
file_idx = 1000
file_idx = 2000
file_idx = 3000
file_idx = 4000

Now it had been doing "something" for 1.5 hours but I can't seem to figure out what. My machine is not unresponsive. What is it supposed to do next?

P.S. I am noob in ML, object detection, sorry if this query sounds dumb.

Problem with inference

Hi, first of all I would like to thank you for this amazing work. However, I have a problem when inferring:

python test.py --data_path=C:\Users\Brani\OneDrive\Bureau\context-based-parking-slot-detect master\train\image\180109_01_000321.jpg

Traceback (most recent call):
File "test.py", line 33, in
pcr_train.evaluate(os.path.join(args.data_path, "test"), args.pcr_test_weight)
File "C:\Users\Brani\OneDrive\Bureau\context-based-parking-slot-detect-master\parking_context_recognizer\train.py", line 78, in evaluate
type_predict, angle_predict = test_model.predict(image, steps = int(len(filenames)/BATCH_SIZE + 1))
ValueError : not enough values to decompress (expected 2, got 0)

Thanks in advance for your answer.

The model of parking_slot_detector missing.

Hello, The work is great, but the module of parking slot detector is missing. Could you upload the module, Thanks.

Some questions about the ps2.0_convert dataset?

Thanks a lot for your working.
I have some questions about the converted ps2.0 dataset. In ps2.0, a plot only have two mark-points , how did you generate the left two points and where is the occupied attribute from?

Thanks again and look forward to your reply!
nice day

tensorflow version

Thanks for the Amazing work!

However, what tensorflow version did you use?:)
Or could you provide requirements.txt file?

hello，is there an inference script here?

The weights is missing (with downloaded weight files)

Hello. Thank you for your work. But the weight is missing that is downloaded weight. Could your upload the weight.

TypeError: __init__() got an unexpected keyword argument 'save_freq'

when I train dataset:
python train.py --data_path=$my_data_path

I meat the problem:
cp_callback = tf.keras.callbacks.ModelCheckpoint(checkpoint_path, verbose=1, save_weights_only=True,save_freq='epoch', period=10)
TypeError: init() got an unexpected keyword argument 'save_freq'

Weight_pcr files are emty

i hace succefully prepared the data , but when i run python train.py , i got no issues but the generated weight_pcr and weight_psd are emty !!!
what have i doen wrong ??

Inference on single image

How to use the model to visualise the result on a single image without giving any label?

project requirements

Hello dohoseok, I want to test your project in other scenarios, but I encountered some errors when training PSD. Can you provide project requirements? Thanks

Could you please give an explain of the label txt?

One example of label/180109_00_002039.txt :
0
-5
1 106 259 128 541 235 531 211 251

The last eight looks like represents the BOX.
Could please elaborate what the other numbers represent?
Thank you.

how to How to calculate anchor

how to How to calculate anchor? I'm looking forward to your reply