# CRCC: Contrast-Based Robust Cross-Subject and Cross-Site Representation Learning for EEG

arXiv: https://arxiv.org/abs/2602.19138 · HTML: https://mlsanity.com/paper/2602.19138/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 4 of its 15 references that are in the map.
- First posted on arXiv 22 Feb 2026.

## Builds on (most cited first, 4 of 4)

- [Large Brain Model for Learning Generic Representations with Tremendous EEG Data in BCI](https://mlsanity.com/paper/2405.18765/) (2024)
- [Contrastive Learning of Subject-Invariant EEG Representations for Cross-Subject Emotion Recognition](https://mlsanity.com/paper/2109.09559/) (2021)
- [Multi-dataset Joint Pre-training of Emotional EEG Enables Generalizable Affective Computing](https://mlsanity.com/paper/2510.22197/) (2025)
- [Differences in Neurovascular Coupling in Patients with Major Depressive Disorder: Evidence from Simultaneous Resting-State EEG-fNIRS](https://mlsanity.com/paper/2506.11634/) (2025)

## Built on by (most cited first, 1 of 1)

- [Beyond Augmentation: Score-Guided Pathological Prior for EEG-based Depression Detection](https://mlsanity.com/paper/2606.00180/) (2026)

## Abstract (arXiv)

EEG-based neural decoding models often fail to generalize across acquisition sites due to structured, site-dependent biases implicitly exploited during training. We reformulate cross-site clinical EEG learning as a bias-factorized generalization problem, in which domain shifts arise from multiple interacting sources. We identify three fundamental bias factors and propose a general training framework that mitigates their influence through data standardization and representation-level constraints. We construct a standardized multi-site EEG benchmark for Major Depressive Disorder and introduce CRCC, a two-stage training paradigm combining encoder-decoder pretraining with joint fine-tuning via cross-subject/site contrastive learning and site-adversarial optimization. CRCC consistently outperforms state-of-the-art baselines and achieves a 10.7 percentage-point improvement in balanced accuracy under strict zero-shot site transfer, demonstrating robust generalization to unseen environments.