Metraipolis › archive › dc_bb0fbb7d6184

Selective Backpropagation for Efficient Few-Shot Class-Incremental Learning — arXiv:2610.04003

Source
arXiv
Original
https://arxiv.org/abs/2610.04003
License (as archived)
CC0-1.0
Archived
2026-10-06T05:25:54.658Z
SHA-256 of the archived text
31f821e9ad2eb4df997f0637975b7f6d401c47ec972b39c1c3251d0647efaf2e
Stable URL of this page
https://metraipolis.com/documento/dc_bb0fbb7d6184

How to cite

Selective Backpropagation for Efficient Few-Shot Class-Incremental Learning — arXiv:2610.04003 — arXiv, https://arxiv.org/abs/2610.04003 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bb0fbb7d6184 (Metraipolis archive dc_bb0fbb7d6184, sha256 31f821e9ad2eb4df997f0637975b7f6d401c47ec972b39c1c3251d0647efaf2e)
BibTeX
@misc{dcbb0fbb7d6184,
  title = {Selective Backpropagation for Efficient Few-Shot Class-Incremental Learning — arXiv:2610.04003},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.04003},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bb0fbb7d6184; sha256 31f821e9ad2eb4df997f0637975b7f6d401c47ec972b39c1c3251d0647efaf2e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.04003. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Few-Shot Class-Incremental Learning (FSCIL) requires models to continuously learn new classes from limited samples while retaining prior knowledge, under strict constraints on compute and memory. Existing approaches lie along a difficult trade-off: simple fine-tuning is computationally efficient but suffers from catastrophic forgetting, replay-based methods mitigate forgetting at the cost of substantial compute and memory, and exemplar-free methods often reduce forgetting by freezing most of the backbone, improving efficiency at the expense of adaptability. We propose Selective Backpropagation (SBP), a deterministic parameter budgeting framework that bridges this gap. SBP restricts gradient updates to a pre-allocated subset of network parameters, freezing past knowledge and preserving unbiased capacity for future learning, enabling rapid adaptation without costly mask optimization. We show that SBP achieves strong performance across standard FSCIL benchmarks while requiring training time close to that of naive fine-tuning and substantially lower training time than prior SOTA methods. Crucially, our experiments expose a limitation of standard FSCIL evaluation: performance on short, distribution-consistent benchmarks does not necessarily predict behavior under distribution shift or over substantially longer learning horizons. We therefore evaluate FSCIL methods in cross-domain settings and over an 80-session ImageNet-1K stream. SBP remains strong across these regimes while maintaining low training cost, providing a favorable stability-plasticity-efficiency trade-off. Our code is available at this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2610.04003 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-06T05:24:05.768Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).