Q-DIVER: Integrated Quantum Transfer Learning and Differentiable Quantum Architecture Search with EEG Data

2026-03-30

AI Summary

This work addresses the limitations of existing quantum circuits in deep learning, which often rely on heuristic designs and lack adaptability to specific task requirements. To overcome this, the authors propose an end-to-end quantum transfer learning framework that integrates a pretrained EEG encoder, DIVER-1, with a differentiable quantum classifier. Notably, they introduce differentiable quantum architecture search for the first time to automatically discover optimal quantum circuit structures. The resulting model achieves a test F1 score of 63.49% on the PhysioNet motor imagery dataset while drastically reducing parameter count—the task-specific head contains only approximately 1/50th the parameters of a classical multilayer perceptron (2.10M vs. 105.02M)—demonstrating an efficient and adaptive approach to quantum neural network design.

Abstract

Integrating quantum circuits into deep learning pipelines remains challenging due to heuristic design limitations. We propose Q-DIVER, a hybrid framework combining a large-scale pretrained EEG encoder (DIVER-1) with a differentiable quantum classifier. Unlike fixed-ansatz approaches, we employ Differentiable Quantum Architecture Search to autonomously discover task-optimal circuit topologies during end-to-end fine-tuning. On the PhysioNet Motor Imagery dataset, our quantum classifier achieves predictive performance comparable to classical multi-layer perceptrons (Test F1: 63.49\%) while using approximately \textbf{50$\times$ fewer task-specific head parameters} (2.10M vs. 105.02M). These results validate quantum transfer learning as a parameter-efficient strategy for high-dimensional biological signal processing.

Problem

• quantum transfer learning
• differentiable quantum architecture search
• EEG data
• parameter efficiency
• quantum-classical hybrid models

Innovation

• Quantum Transfer Learning
• Differentiable Quantum Architecture Search
• Quantum Neural Networks
• EEG Signal Processing
• Parameter Efficiency