[GH-ISSUE #155] Is there away where CLIP gives use multiple outputs like Multiple Object Detection #133

Closed
opened 2026-10-31 21:11:33 +00:00 by bsakaguchi · 1 comment
Owner

Originally created by @Srkshaikh5 on GitHub (Sep 23, 2021).
Original GitHub issue: https://github.com/openai/CLIP/issues/155

Hi,

Can You Help me, where If we ask the CLIP irrelevant questions with Image and it still tries to fit the score even if the class is not there, because of the Softmax layer used.
Or
Is there away where CLIP gives use multiple outputs like if we ask for multiple objects in the image and it gives score for each object separately like we do in object detection.

Originally created by @Srkshaikh5 on GitHub (Sep 23, 2021). Original GitHub issue: https://github.com/openai/CLIP/issues/155 Hi, Can You Help me, where If we ask the CLIP irrelevant questions with Image and it still tries to fit the score even if the class is not there, because of the Softmax layer used. Or Is there away where CLIP gives use multiple outputs like if we ask for multiple objects in the image and it gives score for each object separately like we do in object detection.
Author
Owner

@jongwook commented on GitHub (Sep 24, 2021):

See #82 for related discussions.

<!-- gh-comment-id:926286294 --> @jongwook commented on GitHub (Sep 24, 2021): See #82 for related discussions.
Sign in to join this conversation.
No labels
pull-request
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
openai/CLIP#133
No description provided.