Jul 2026
Spectral-Aware Analytic Class-Incremental Learning for Long-Tailed Distributions
Geometry-Spectral Rectification (GSR), a theoretically grounded framework that treats long-tailed learning as a spectral regularization problem, is proposed, offering a superior trade-off between computational efficiency and robust generalization in long-tailed settings.
Q. Tran, Ngoc-Hai Nguyen, Quan Dao et al.
· arXiv.org · 0 citations