ADP: Adaptive Discriminative Prototypes for Cross-Domain Few-Shot Object Detection
Electronics · 28.10.2026 · 15 (21) · 4991
DOI: 10.3390/electronics15214991 ↗Kurz gefasst
Fachartikel in der Zeitschrift Electronics, online erschienen am 28.10.2026. Verfasst von Saniya M. Deshmukh und 2 weiteren Autor:innen. Die Arbeit nennt 33 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: FCT-Fundação para a Ciência e a Tecnologia, European Union through the European Regional Development Fund, Instituto de Telecomunicações.
Abstract (Originalsprache)
Cross-domain few-shot object detection (CD-FSOD) presents a fundamental challenge: a detector must recognise novel object categories from as few as one to ten labelled examples per class, frequently in visual domains that differ substantially from the training distribution. Leading methods, such as CD-ViTO, address this by constructing per-class prototype representations through Sinkhorn optimal transport clustering of DINOv2 patch-token features. When categories share fine-grained visual texture, however, the resulting prototypes can overlap in feature space, which degrades nearest-neighbour assignment and reduces detection accuracy. To address this problem, we propose ADP (Adaptive Discriminative Prototypes), a training-free post-processing step that increases inter-class prototype separation through difficulty-conditioned anchored repulsion. ADP computes a per-support-set difficulty score δ to quantify inter-class prototype overlap. When δ is below a low threshold, the original Sinkhorn prototypes are returned unchanged. Otherwise, a repulsion strength and margin proportional to δ are applied through a short anchor-regularised gradient step, keeping the refined prototypes tied to the support evidence. On the NTIRE 2026 splits, every measured δ exceeds that threshold, so this deactivation branch is not exercised and its effect is unverified. On the NTIRE 2026 CD-FSOD benchmark, fixed-strength repulsion has been shown to yield large gains on confused domains while degrading performance on well-separated ones. ADP records a higher mAP than the Sinkhorn baseline in 13 of 18 conditions. The average change is +0.19 mAP at 1-shot, +0.03 at 5-shot, and −0.07 at 10-shot. Each number comes from the challenge’s fixed support file, so these changes are directional evidence from a single deterministic run. Relative to fixed-strength repulsion, ADP avoids that variant’s largest regressions, while the average change itself remains small.
Förderung
- FCT-Fundação para a Ciência e a Tecnologia
- European Union through the European Regional Development Fund
- Instituto de Telecomunicações
- AcornSelectAi-Acorn Kernel Selection System Using Artificial Intelligence
Zitieren
APA
Deshmukh, S. M., Hambarde, K. A., & Proença, H. (2026). ADP: Adaptive Discriminative Prototypes for Cross-Domain Few-Shot Object Detection. Electronics, 15(21), 4991. https://doi.org/10.3390/electronics15214991
Vancouver
Deshmukh SM, Hambarde KA, Proença H. ADP: Adaptive Discriminative Prototypes for Cross-Domain Few-Shot Object Detection. Electronics. 2026;15(21):4991. doi:10.3390/electronics15214991
BibTeX
@article{Deshmukh2026,
author = {Deshmukh, Saniya M. and Hambarde, Kailash A. and Proença, Hugo},
title = {ADP: Adaptive Discriminative Prototypes for Cross-Domain Few-Shot Object Detection},
journal = {Electronics},
year = {2026},
volume = {15},
number = {21},
pages = {4991},
doi = {10.3390/electronics15214991}
}
EUVIMED