A dual-stage energy-based active learning method filters unknown classes with one EBM and scores known samples with another to improve selection efficiency in open-set object classification.
In: Proceedings of the International Conference on Machine Learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Energy-Based Open-Set Active Learning for Object Classification
A dual-stage energy-based active learning method filters unknown classes with one EBM and scores known samples with another to improve selection efficiency in open-set object classification.