Multi-channel convolutional neural quantum embedding

2025-09-26

AI Summary

Classical multi-channel data suffers from low embedding efficiency in quantum supervised learning, constrained by the requirement of completely positive trace-preserving (CPTP) maps. Method: This paper proposes a non-CPTP multi-channel convolutional neural quantum embedding framework, integrating classical convolutional feature extraction with parameterized quantum circuits in a hybrid architecture. It overcomes channel limitations inherent in standard quantum circuit models via channel-specific quantum state encoding and learnable embedding weights. Contribution/Results: Experiments on CIFAR-10 and Tiny ImageNet demonstrate significant improvements in classification accuracy (+3.2–5.7%) and generalization performance. Theoretical analysis establishes the validity and superiority of non-CPTP embedding for high-dimensional feature representation. This work introduces a novel paradigm for efficient classical data embedding in quantum machine learning.

Technology Category

• Machine Learning: Quantum Machine Learning
• Computer Vision: Learning & Optimization for CV
• Search and Optimization: Non-convex Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Security and Privacy: Data transparency and provenance
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Classification using variational quantum circuits is a promising frontier in quantum machine learning. Quantum supervised learning (QSL) applied to classical data using variational quantum circuits involves embedding the data into a quantum Hilbert space and optimizing the circuit parameters to train the measurement process. In this context, the efficacy of QSL is inherently influenced by the selection of quantum embedding. In this study, we introduce a classical-quantum hybrid approach for optimizing quantum embedding beyond the limitations of the standard circuit model of quantum computation (i.e., completely positive and trace-preserving maps) for general multi-channel data. We benchmark the performance of various models in our framework using the CIFAR-10 and Tiny ImageNet datasets and provide theoretical analyses that guide model design and optimization.

Problem

• Optimizing quantum embedding for multi-channel data
• Enhancing quantum supervised learning classification performance
• Benchmarking hybrid classical-quantum models on image datasets

Innovation

• Multi-channel convolutional neural quantum embedding
• Classical-quantum hybrid quantum embedding optimization
• Variational quantum circuits for data classification