GiPL uses iterative pseudo-label self-training on support sets plus generative augmentation from VLMs to improve CD-FSOD performance on RUOD, CARPK, and CarDD under 1/5/10-shot regimes.
Interpretable cross-domain few-shot learning with rectified target-domain local alignment.arXiv preprint arXiv:2603.17655, 2026
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
-
GiPL: Generative augmented iterative Pseudo-Labeling for Cross-Domain Few-Shot Object Detection
GiPL uses iterative pseudo-label self-training on support sets plus generative augmentation from VLMs to improve CD-FSOD performance on RUOD, CARPK, and CarDD under 1/5/10-shot regimes.