aras62 / sf-gru Goto Github PK

Hi, thanks for impressive work! I want to only run the test (using the pretrained you have provided) but when I run, I get this error:

File "/home/SF-GRU/test.py", line 9, in
beh_seq_test = imdb.generate_data_trajectory_sequence('test', **data_opts)
File "/home/SF-GRU/pie_data.py", line 891, in generate_data_trajectory_sequence
sequence_data = self._get_crossing(image_set, annot_database, **params)
File "/home/SF-GRU/pie_data.py", line 1003, in _get_crossing
set_ids, _pids = self._get_data_ids(image_set, params)
File "/home/SF-GRU/pie_data.py", line 791, in _get_data_ids
_pids = self._get_random_pedestrian_ids(image_set, **params['random_params'])
File "/home/SF-GRU/pie_data.py", line 724, in _get_random_pedestrian_ids
train_samples, test_samples = train_test_split(ped_ids, train_size=ratios[0])
File "/home/SF-GRU/pedintent/lib/python3.5/site-packages/sklearn/model_selection/_split.py", line 2122, in train_test_split
default_test_size=0.25)
File "/home/SF-GRU/pedintent/lib/python3.5/site-packages/sklearn/model_selection/_split.py", line 1805, in _validate_shuffle_split

ValueError: With n_samples=0, test_size=None and train_size=0.5, the resulting train set will be empty. Adjust any of the aforementioned parameters.

I have put the test annotations in the following format:
SF-GRU
pie_dataset
annotations
set01
video_0001_annt.xml
video_0002_annt.xml
video_0003_annt.xml
video_0004_annt.xml

Do you know what the issue is? If you could be provide more detail about the test and train file (or sample.py file in the repo), that will be very helpful. Thanks!

Hi Amir,

I was trying to run the test script. However it only generated 656 test samples, not 3185 that you mentioned in your paper. Could you please let me know how you get that 3185 test samples?

I added a break point here and then find the length of test_data['test'][0][0] is 656:

Thank you,
Brian

Hi,
Thanks for this wonderful works, it brings a lot of idea for my own research.
I wonder how to test vary ttc prediction behavior after I trained a model.
I just set
"min_track_size" to be observation_time + ttc in test_script.py
However, no matter how I set this value, the output of accuracy is still high (above 0.85 up)
How is this possible when it has such accuracy for about 10 sec event prediction?
Sincerely

One thing you might want to check first is to make sure to delete the dataset cache file and run it again. If you change the sets in _get_image_set_ids it might have an issue with the cached file. Second thing is that what is the dimension of d['act']? it should be three dimensions as [num_samples, seq_length, 1] and after d['act'] = d['act'][:,0,:] should be [num_samples,1]. If you changed anything in the get_sequence, it might affect that

Originally posted by @aras62 in #1 (comment)

Hello authors, firstly thank you for your contribution. I just wanted to take your help in fixing this issue. I have taken the PIE dataset from the repository and having this index out of dimensions error. I am also attaching screenshot of the output screen when running on colab. Moreover as you suggested to check the dimensions. I am getting dimensions of all the elements of dictionary d as (num_samples, ). So does this dataset require some preprocessing? as I am also getting some visibledeprecation warning.

Also I am not changing anything in get_sequence nor I am running on some part of dataset. I am running on complete 6 set of videos. Kindly help