talmolab/sleap

Still receive `KeyError` even after re-encoding

rhansen123 opened this issue · 6 comments

Bug description

I have been trying to run batches of videos through SLEAP inference, but SLEAP has had issues with a few videos. The error in the command terminal is attached to this post, but it seems to have an issue with a specific frame in a video and not finish running an inference. I reencoded the videos through Python as suggested, but the issue still occurs.

Expected behaviour

SLEAP should be able to run inference on the videos.

Actual behaviour

It stops at a few different videos with an error that it cannot load a frame.
Capture

  • OS:
  • Version(s):
Environment packages
# paste output of `pip freeze` or `conda list` here
Logs
# paste relevant logs here, if any

Screenshots

How to reproduce

  1. Go to '...'
  2. Click on '....'
  3. Scroll down to '....'
  4. See error

Hey @rhansen123,

Glad that you tried the re-encoding. That usually does the trick. Not glad it didn't work. The useful piece of information here would be what happened after you tried re-encoding your video? Did the terminal say anything? I expect it should have since it sounds like the re-encoding failed and that's why things are not working.

Thanks,
Liezl

Closing this in favor of #1711.

@rhansen123: This will be fixed in the next release, but feel free to install from source with the latest version from the develop branch if you're blocked.

@talmo Came across this issue as I was running into the same issue (I resolved it in the end, but wasn't that easy to sort out). When do you expect the next release to come out? :-)

Haha, I'll take as a "let's hear what the other folks think" 😂 Looking forward to it nevertheless!