Do sEMG Foundation Models Improve Cross-Subject Generalisation?
Abstract
sEMG Foundation Models aim to learn general purpose representations over diverse subjects and recording conditions that boost performance on a variety of downstream tasks. However, existing evaluations make it difficult to discern whether pretraining alone is sufficient to combat the cross-subject overfitting which makes machine learning on sEMG so challenging. We conduct a systematic evaluation of three sEMG foundation models across three gesture classification datasets under both intra-session and inter-subject splits, using three fine-tuning strategies designed to isolate the contribution of pretraining. We find that pretraining boosts performance when train and test samples are drawn from the same recording session, but this benefit largely disappears under inter-subject evaluation. In low subject count datasets, pretraining provides little to no benefit and classical ML baselines match or exceed foundation model performance.
Related
- Harmonised benchmarking of foundation models for single-cell and spatial transcriptomics reveals context-dependent generalisation
- Longitudinal Subject Pairing in Cross-Sectional Neonatal Data Reveals Asynchronous Structural and Functional Brain Maturation
- Latent Prediction Yields Cross-Subject Alignable Representations in Sensory-Evoked fMRI
- Leakage-Audited Benchmarking Reveals Limited Evidence for Cross-Subject Auditory-Evoked EEG Vowel Perception Decoding
- When Model Complexity Matters in Cross-Subject Naturalistic fMRI Prediction
BibTeX
@misc{raftery2026semg,
title = {{Do sEMG Foundation Models Improve Cross-Subject Generalisation?}},
author = {William Raftery and Dario Farina},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.08199},
eprint = {2611.08199},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.08199},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Do sEMG Foundation Models Improve Cross-Subject Generalisation?
AU - William Raftery
AU - Dario Farina
PY - 2026
DA - 2026/11/05
PB - arXiv
DO - 10.48550/arxiv.2611.08199
UR - https://arxiv.org/abs/2611.08199
AB - sEMG Foundation Models aim to learn general purpose representations over diverse subjects and recording conditions that boost performance on a variety of downstream tasks. However, existing evaluations make it difficult to discern whether pretraining alone is sufficient to combat the cross-subject overfitting which makes machine learning on sEMG so challenging. We conduct a systematic evaluation of three sEMG foundation models across three gesture classification datasets under both intra-session and inter-subject splits, using three fine-tuning strategies designed to isolate the contribution of pretraining. We find that pretraining boosts performance when train and test samples are drawn from the same recording session, but this benefit largely disappears under inter-subject evaluation. In low subject count datasets, pretraining provides little to no benefit and classical ML baselines match or exceed foundation model performance.
ER -