lightweight-deep-learning

Ensemble Deep Random Vector Functional Link with Skip Connections (edRVFL-SC) No GPU required • 100× faster training

The official repository for "Calibration-Aware Lightweight Model Selection for Trustworthy Skin Lesion Classification" accepted at the MIWAI2026.

LumiNet is a research-focused compact single-stage CNN for joint low-light image enhancement and denoising, designed to restore illumination, suppress noise, and preserve structural detail in a single forward pass.

MSNet / LS-Net: Multi-Scale Representation Networks for Time Series Classification Reference PyTorch implementation of MSNet and LS-Net, two lightweight multi-scale CNN architectures for time series classification. The repository provides model implementations and runnable demos on UCR datasets.

Five-fold evaluation of lightweight deep-learning models for referable diabetic-retinopathy screening.