haozhiqi / rpin Goto Github PK

Hi,

Thank you very much for sharing your code! I just see that the ROIs locations are input to the model together with the first 4 frames, however, I am sure how are the ROIs feature are cat over time? How we can know that two ROIs belong to the same object? Are the ROIs annotated with an instance ID?

Best,

Hi,

Thanks for sharing your work. I am wondering how did you get the dynamic objects ROIs. It seems like from PHYRE API, they output info about both dynamic and static objects in random order. Did you manually label the dynamic objects ROIs?

when i run simb/prepare_billiard.py:

OSError: [Errno 5] Can't read data (file read failed: time = Sun Dec 4 22:20:06 2022
, filename = 'Dataloader/simb/test.hkl', file descriptor = 4, errno = 5, error message = 'Input/output error', buf = 0x564eae5a1151, total read size = 907, bytes this sub-read = 907, bytes actually read = 18446744073709551615, offset = 0)

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
File "Dataloader/simb/prepare_billiard.py", line 73, in
main()
File "Dataloader/simb/prepare_billiard.py", line 31, in main
src_data = hickle.load(src_file)
File "/remote-home/xhl/anaconda3/envs/gsgn/lib/python3.7/site-packages/hickle/hickle.py", line 387, in load
raise ValueError("Provided argument 'file_obj' does not appear to be a valid hickle file! (%s)" % (error),error) from error
ValueError: ("Provided argument 'file_obj' does not appear to be a valid hickle file! ([Errno 5] Can't read data (file read failed: time = Sun Dec 4 22:20:06 2022\n, filename = 'Dataloader/simb/test.hkl', file descriptor = 4, errno = 5, error message = 'Input/output error', buf = 0x564eae5a1151, total read size = 907, bytes this sub-read = 907, bytes actually read = 18446744073709551615, offset = 0))", OSError(5, "Can't read data (file read failed: time = Sun Dec 4 22:20:06 2022\n, filename = 'Dataloader/simb/test.hkl', file descriptor = 4, errno = 5, error message = 'Input/output error', buf = 0x564eae5a1151, total read size = 907, bytes this sub-read = 907, bytes actually read = 18446744073709551615, offset = 0)"))

Hi Haozhi,

Very nice work! I am currently following your work. Could you share the pre-training models or training codes of other baselines in Tab. 1 (VIN, OM, CVP, IN)?

Thanks,
Mingyu