ThinkNet finds compact brain signal decoders that work for new users
ThinkNet: Compact Architecture Selection and Validation-Gated Ensembles for Subject-Independent MI-EEG Decoding
Human-Computer InteractionArtificial IntelligenceMachine Learning
Summary
Brain-computer interfaces can help people by decoding brain signals linked to imagined movements. But creating decoders that work well for new people without much training data is tough. The authors created ThinkNet, a system that carefully selects small, efficient models and tests them without using data from the new person. Their approach finds compact models that perform better on new users than bigger ones and combines multiple models in a way that improves accuracy. Still, their work shows challenges remain in reliably choosing the best models when the brain signals vary across individuals.
What this means in practice
- •For assistive technology developers: Develop portable brain signal decoders that generalize well to new users with limited calibration data for motor-imagery applications.
- •For embedded system engineers: Deploy compact and fast brain-signal decoding models on low-power devices, improving real-time performance for assistive interfaces.
Authors
Abdul Basit, Saim Rehman, Muhammad Shafique
Abstract
Practical assistive and rehabilitative brain--computer interfaces require subject-independent motor-imagery EEG (MI-EEG) decoders that generalize to new users under limited target-user data and constrained compute. However, held-out-subject performance can be overstated when test-subject information influences preprocessing, model selection, or ensemble selection. We present \textit{ThinkNet}, a validation-controlled framework that combines train-only normalization, validation-guided evolutionary search, and validation-gated inference to identify compact decoders and inference policies for held-out subjects. We evaluate four-class BCI Competition IV-2a (session T) decoding with nine Leave-One-Subject-Out (LOSO) folds, three seeds, seven fixed decoder entries, and a broader search over ten representative decoder families; the held-out subject is never used for normalization, hyperparameter, architecture, or ensemble-policy selection. In the fixed benchmark, the validation-selected compact decoder achieved 44.35$\pm$15.41\% accuracy with 4.9K parameters, 19 KB FP32 weights, and 0.99 ms batch-1 Orin CUDA inference. Across the broader search, compact models ($\leq$25K parameters) achieved higher mean held-out accuracy than mid-size and large alternatives after selected retraining (40.10\% vs. 35.09\% and 34.78\%). Validation-gated ensembling improved over validation-selected single-model inference, reaching 43.98$\pm$16.25\% in the fixed benchmark and 43.31$\pm$15.88\% for the compact six-family ensemble. A non-deployable oracle analysis revealed a 6.1-point family-selection gap and near-zero validation--test correlation, showing that validation reliability remains a key bottleneck under subject shift. Thus, ThinkNet is a validation-controlled framework for compact MI-EEG model and inference-policy selection, rather than a single-architecture benchmark.