CalexNet: Soft Cascade-Aligned Training and Calibration for Lightweight Early-Exit Branches
Electronics · 16.05.2026 · 15 (10) · 2149
DOI: 10.3390/electronics15102149 ↗Kurz gefasst
Fachartikel in der Zeitschrift Electronics, online erschienen am 16.05.2026. Verfasst von Yehudit Aperstein und 1 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 43 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Early-exit cascades over a frozen convolutional backbone enable adaptive inference but suffer from three sources of train–inference mismatch: branches train on samples they will never see at inference; their per-class precision thresholds are calibrated on the wrong distribution; the standard cross-entropy target on backbone argmax labels discards the backbone’s uncertainty signal. We close all three gaps with CalexNet (cascade-aligned early exits), a training-recipe-only modification: branches train under continuously weighted importance sampling that matches the cascade-survivor distribution; per-class precision thresholds are calibrated on the actual cascade-survivor subset of the validation set; the classification head is trained against the backbone’s full softmax via a temperature-scaled KL objective. Combined with an augmented prototype-pooling branch head, CalexNet is evaluated on ResNet18 and ResNet50 backbones across CIFAR-100 (20-supe-class coarse, the harder primary setting) and CINIC-10 (10-class, the easier cross-validation counterpart). On the accuracy–FLOPs Pareto frontier, CalexNet matches or exceeds three published baselines (PTEEnet, ZTW, BoostNet) and a within-paper “no-alignment, no-KD” reference. The largest gains appear in the practically relevant 30–70% FLOPs-reduction regime and show consistent trends across n=3 training seeds. CalexNet requires no inference-time architectural change and is a drop-in for any frozen-backbone early-exit cascade.
Zitieren
APA
Aperstein, Y., & Apartsin, A. (2026). CalexNet: Soft Cascade-Aligned Training and Calibration for Lightweight Early-Exit Branches. Electronics, 15(10), 2149. https://doi.org/10.3390/electronics15102149
Vancouver
Aperstein Y, Apartsin A. CalexNet: Soft Cascade-Aligned Training and Calibration for Lightweight Early-Exit Branches. Electronics. 2026;15(10):2149. doi:10.3390/electronics15102149
BibTeX
@article{Aperstein2026,
author = {Aperstein, Yehudit and Apartsin, Alexander},
title = {CalexNet: Soft Cascade-Aligned Training and Calibration for Lightweight Early-Exit Branches},
journal = {Electronics},
year = {2026},
volume = {15},
number = {10},
pages = {2149},
doi = {10.3390/electronics15102149}
}
EUVIMED