Metraipolis › archive › dc_6cb84cdf99d0

GaLe: memory-efficient Global Approximate and Local Exact features — arXiv:2609.02689

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.02689
License (as archived)
CC0-1.0
Archived
2026-10-17T05:50:14.178Z
SHA-256 of the archived text
1bc25fec3b4f38439b45bad41856ae828060bad39468fe36b7bec137b2ede86e
Stable URL of this page
https://metraipolis.com/documento/dc_6cb84cdf99d0

How to cite

GaLe: memory-efficient Global Approximate and Local Exact features — arXiv:2609.02689 — arXiv (en), https://arxiv.org/abs/2609.02689 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_6cb84cdf99d0 (Metraipolis archive dc_6cb84cdf99d0, sha256 1bc25fec3b4f38439b45bad41856ae828060bad39468fe36b7bec137b2ede86e)
BibTeX
@misc{dc6cb84cdf99d0,
  title = {GaLe: memory-efficient Global Approximate and Local Exact features — arXiv:2609.02689},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.02689},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_6cb84cdf99d0; sha256 1bc25fec3b4f38439b45bad41856ae828060bad39468fe36b7bec137b2ede86e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.02689. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Embedded devices typically lack the resources of GPU-equipped machines, and existing inference methods suffer from either high computational overhead (patch-based) or accuracy loss (approximation-based). We propose GaLe, a memory-efficient technique that enables the deployment of pretrained networks on constrained devices without retraining. GaLe partitions feature maps into two components: a local exact (Le) representation that preserves fine details and a global approximate (Ga) representation that retains long-range dependencies. Unlike standard tiling, GaLe supports global operations and attention mechanisms found in hybrid CNN-transformer models. Validated on ImageNet, our method matches exact-inference performance while achieving up to 65% speedup and 90% RAM reduction on a Cortex-M33 compared to patch-based inference. We further demonstrate GaLe's versatility across classification, detection, and generation tasks, highlighting its potential as a foundation for resource-efficient architecture design.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-03 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.02689 — 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 2026-10-17T05:25:58.152Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-03?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).