DSAA improves fine-grained open-vocabulary object detection by injecting attribute priors via APA in text embeddings, modulating K/V vectors in BERT, and using an attribute-aware contrastive loss, with gains shown on the FG-OVD benchmark.
The devil is in the fine-grained details: Evaluating open-vocabulary object detectors for fine-grained understanding
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
DSAA: Dual-Stage Attribute Activation for Fine-grained Open Vocabulary Detection
DSAA improves fine-grained open-vocabulary object detection by injecting attribute priors via APA in text embeddings, modulating K/V vectors in BERT, and using an attribute-aware contrastive loss, with gains shown on the FG-OVD benchmark.