jjn123 / fall-detection Goto Github PK

i need to run the code on other similar datasets plz elaborate how to generate .npy files for a dataset that you have used in your code

the opencvinpaint seems didnot work for the URFD dateset but work the SDU FALL

hi! when I run the dae_main_test.py with UR_filled dataset, it casue the error belows. But when I run the thermal dataset, it doesn‘t make a mistake. So can you know what's the problem？ thanks

Traceback (most recent call last):
File "dae_main_test.py", line 22, in
dae_exp.test(raw = False, animate = True)
File "/home/rubin/papercode/Fall-Detection-master/ae_exp.py", line 117, in test
metric = 'AUC', dset = self.dset, raw = raw, animate = animate) #multi_AEmodel.py
File "/home/rubin/papercode/Fall-Detection-master/multi_AEmodel.py", line 73, in get_stats_for_all_vids
vid_total = data_dict[Fall_name]['Data'][:]
File "h5py/_objects.pyx", line 54, in h5py._objects.with_phil.wrapper
File "h5py/_objects.pyx", line 55, in h5py._objects.with_phil.wrapper
File "/home/rubin/anaconda3/envs/tensorflow/lib/python3.7/site-packages/h5py/_hl/group.py", line 264, in getitem
oid = h5o.open(self.id, self._e(name), lapl=self._lapl)
File "h5py/_objects.pyx", line 54, in h5py._objects.with_phil.wrapper
File "h5py/_objects.pyx", line 55, in h5py._objects.with_phil.wrapper
File "h5py/h5o.pyx", line 190, in h5py.h5o.open
KeyError: "Unable to open object (object 'Fall1' doesn't exist)"

Why can't I load the dataset

hi! I can not found the loss fuction in your code,I only see this autoencoder.compile(optimizer='adadelta', loss='mean_squared_error'). So if I want to change the loss fuction,what should I do? Or where the loss fuction is?
thanks!

I have already contacted Dr. Khan by email. Maybe my mail is classified as spam. Can you still provide other ways to get processed data?

Sincerely,

I looked into it and found that there are many types of thermal imaging cameras and depth cameras. Which one did you use?

How come DSTCAE_C3D model replaced by dummy_3d?

    model, model_name, model_type = DSTCAE_C3D(img_width, img_height, win_len)
    model, model_name, model_type = dummy_3d(img_width, img_height, win_len)  --?

1. I would like to ask how to give a label for each picture for a set of falling pictures when preparing the training set?
2. When testing a group of pictures, how to output start frame index and stop frame index of a fall action?

decoded_imgs = self.model.predict(test_data)
RE = MSE(test_data, decoded_imgs)
predicted_classes = threshold(RE, 0.5)

Is the “predicted_classes” the label of the model's final prediction?

Where to get SDUFall dataset?