Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Mask Generation Tasking (adding about.md) #438

Merged
merged 16 commits into from
Jan 26, 2024
Merged

Conversation

bellabf
Copy link
Contributor

@bellabf bellabf commented Jan 15, 2024

Hi,

I wrote an "about.md" for the "Mask generation task".
I saw that @nebulaanish had already opened a PR for the summary, so I paddle back on the "data.ts" in favor of his (thus, the commits).

Thanks!

Copy link
Contributor

@merveenoyan merveenoyan left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks a lot for this PR <3

.vscode/settings.json Outdated Show resolved Hide resolved
packages/tasks/src/tasks/mask-generation/data.ts Outdated Show resolved Hide resolved
packages/tasks/src/tasks/mask-generation/about.md Outdated Show resolved Hide resolved

Generating masks that can to separate background and foreground (or other interesting properties) in an image can be done to facilitate learning, especially in semi- or unsupervised learning.

### Human-in-the-loop
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can you elaborate the use case here? i.e. The name of the use case felt a little ambiguous

packages/tasks/src/tasks/mask-generation/about.md Outdated Show resolved Hide resolved
packages/tasks/src/tasks/mask-generation/about.md Outdated Show resolved Hide resolved
packages/tasks/src/tasks/mask-generation/about.md Outdated Show resolved Hide resolved
packages/tasks/src/tasks/mask-generation/about.md Outdated Show resolved Hide resolved
packages/tasks/src/tasks/mask-generation/about.md Outdated Show resolved Hide resolved

### Pre-processing

Generating masks that can to separate background and foreground (or other interesting properties) in an image can be done to facilitate learning, especially in semi- or unsupervised learning.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this might actually be used to train segmentation models

@bellabf
Copy link
Contributor Author

bellabf commented Jan 21, 2024

Thanks a lot for this PR <3

Hi, Merve! Thank you so much for the review. I had misunderstood the task variation part and I fixed it now following your suggestion. I tried to create a better naming for the second use case you mentioned, but I am a bit out of ideas. In more plain terms, the idea is:

  • pre-processing= I know that the task only concern this part of the image, so I will use the mask to tell the model look here
  • masked image modelling= I'll train the model to reconstruct the masked part of the image

Thanks again <3

@merveenoyan
Copy link
Contributor

@bellabf can you run the following in this repository?

sudo corepack enable
pnpm install

pnpm -r format
pnpm -r lint

and some files in this PR should change. Can you add, commit and push them?

@merveenoyan merveenoyan merged commit 2ca74cc into huggingface:main Jan 26, 2024
2 checks passed
@merveenoyan
Copy link
Contributor

@bellabf @nebulaanish if you could give me your HF handles I can add them at the end of the published page for credits

@bellabf
Copy link
Contributor Author

bellabf commented Jan 26, 2024

@bellabf @nebulaanish if you could give me your HF handles I can add them at the end of the published page for credits

Hi, Merve! Thanks, it is bisnotforbella

@nebulaanish
Copy link
Contributor

nebulaanish commented Jan 27, 2024

Hi, my handle is nebulaanish. Thank you.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants