SOURCE-LINKED INTELLIGENCE

Pretraining and Distillation Matter More Than Architecture Family for Label-Free Single-Cell Classification

arXiv · AI, language, vision and robotics · article · Sep 9, 2026 · UTC

Choosing a deep learning architecture for label-free single-cell classification remains an open question, with microscopy benchmarks reporting conflicting conclusions about CNNs versus transformers. We present a controlled benchmark on LIVECell phase-contrast microscopy data using source-image-disjoint train/validation/test splits to prevent parent-image leakage and matched optimisation, augmentation, and evaluation protocols across EfficientNet, Vision Transformer (ViT), and EVA-02 models. This allows the effects of architecture, pretraining, fine-tuning, tokenisation, and distillation to be

Read original source ↗ Open in workspace

recordType

paper

region

Global