A concept bottleneck model that derives its concepts from segmentation and detection foundation models instead of text, achieving competitive accuracy and better out-of-distribution robustness with only 50 images per class for concept generation.
End-to-end object detection with transformers
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
DCBM: Data-Efficient Visual Concept Bottleneck Models
A concept bottleneck model that derives its concepts from segmentation and detection foundation models instead of text, achieving competitive accuracy and better out-of-distribution robustness with only 50 images per class for concept generation.