epic-kitchens / c1-action-recognition Goto Github PK

Thanks for providing these helpful resources.
I found "We provide an evaluation script to compute the metrics we report in the paper on the validation set." part in the readme.
Can you also share the script and exemplary command to compute the metrics on the test (not validation) set?

I tried to use EPIC_100_test_timestamps.pkl to replace EPIC_100_validation.pkl, but got the following error:

Traceback (most recent call last):
File "evaluate.py", line 127, in
main(parser.parse_args())
File "evaluate.py", line 75, in main
labels = add_action_class_column(labels)
File "evaluate.py", line 66, in add_action_class_column
groundtruth_df["verb_class"], groundtruth_df["noun_class"]
File "/mnt/miniconda/envs/epic-models/lib/python3.8/site-packages/pandas/core/frame.py", line 2902, in getitem
indexer = self.columns.get_loc(key)
File "/mnt/miniconda/envs/epic-models/lib/python3.8/site-packages/pandas/core/indexes/base.py", line 2891, in get_loc
raise KeyError(key) from err
KeyError: 'verb_class'

Hi,

When I run create_submission.py I am getting the error -

AttributeError: 'Namespace' object has no attribute 'scores_pt'

I am using the TSM model for testing

Thanks
Abhinav

Hello, I am a graduate student in computer science and am currently exploring the field of egocentric action recognition, and your Epic-kitchens dataset has given me great space to think and research motivation. Now I want to test my work and would like to ask if I can still submit a test file for test in EPIC-KITCHENS-100 Action Recognition competition in CodaLab website? The decision is pending after the participation module is registered. Look forward to receiving your reply. Thank you.