-
Notifications
You must be signed in to change notification settings - Fork 145
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
update zsod-data.ts #443
base: main
Are you sure you want to change the base?
update zsod-data.ts #443
Conversation
There was a problem hiding this 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! BTW @GuichardVictor is contributing the same page in this PR: #435 maybe you could align with this changes and contribute on top of it?
@@ -0,0 +1,15 @@ | |||
## Use Cases |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think if you're not contributing content to this page you can remove the file itself since placeholder pretty much says the same thing
id: "imagenet-1k" | ||
}, | ||
{ | ||
description: "Microsoft COCO, a large-scale dataset of object instance segmentation, aims to advance object recognition by placing it in the context of scene understanding. It comprises 328k images with 2.5 million labeled instances of 91 common objects easily recognizable by a 4-year-old. ", |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
In1k is enough for now, we can remove coco
demo: { | ||
inputs: [ | ||
{ | ||
filename: "zero-sh-obj-detection_1.png" |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can you upload this image by opening a PR to here? https://huggingface.co/datasets/huggingfacejs/tasks
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
also would be nice to stick to general convention of how we name files (you can take a look at other task pages for example)
type: "chart", | ||
data: [ | ||
{ | ||
label: "human face", |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Zero shot object detection doesn't return a chart but directly a class probability with bounding box, you can take a look at the example task page https://huggingface.co/tasks/object-detection (and it's files in this repository)
], | ||
models: [ | ||
{ | ||
description: "OWL-ViT is a zero-shot text-conditioned object detection model that uses CLIP as its multi-modal backbone to detect objects in images based on textual descriptions. It can handle multiple text queries per image and has been trained on standard detection datasets using a bipartite matching loss." |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We tend to keep the descriptions shorter
@leandstvh can you solve merge conflicts? |
added details to data.ts file for zero-shot object detection