hkchengrex/Cutie

Simple cli inference script

KaareLJensen opened this issue · 3 comments

KaareLJensen commented

Great work !

Will you provide a simple command line interface to process a video/images of video given object mask(s)?
Maybe something like done for XMem++: https://github.com/max810/XMem2#use-xmem-command-line-and-python-interface

I can see that eval_vos.py provides for something like this using dataset=generic, but this script seems more to be for big batch processing of datasets.

hkchengrex commented

eval_vos.py works for a single video (putting just one video in the dataset).
The GUI also works for single video testing.
Our Colab demo shows how to use the codebase for evaluating on a single video.

I agree that having a command line interface for single video evaluation (with potentially multiple permanent memory frames) can be useful. Pull requests are welcomed.

KaareLJensen commented

Thanks for your response.
If I succeed in creating a CLI script for single video evaluation, I will try to create a PR.
I will return to the Colab demo for inspiration ;-)

  • 👍1
KaareLJensen commented

Now added a PR with new CLI script to run on video file #16
Hope it can be used 🙂