Uploaded October 2025 | Updated September 2026, 2 weeks ago
Quantum Machine Learning for Cybersecurity: Malware Detection with Photonic QCNNs:
Quantum Machine Learning (QML) is widely regarded as one of the most promising near-term applications of quantum computing. In this work, we introduce a photonic quantum convolutional neural network (QCNN) architecture designed for classification tasks. Our approach leverages the natural strengths of photonic hardware—including scalability, low energy consumption, and native implementation of linear transformations—to build a noise-resilient QML model. Benchmarking against classical baselines, we find that photonic QCNNs can achieve competitive accuracy. Crucially this benchmark shows that when complexifying the dataset, we have an advanatage for quantum models in the number of needed parameters to reach a given accuracy. To explore concrete applications, we are partnering with Orange to investigate the use of photonic QCNNs for malware detection in cybersecurity. Malware search and classification represent a critical industrial challenge, where fast and energy-efficient pattern recognition can deliver substantial business value. By combining photonic hardware with advanced QML architectures, our collaboration aims to demonstrate how quantum computing can address real-world security threats while maintaining sustainability and scalability.
q2b.qcware.com
Quantum Machine Learning for Cybersecurity: Malware Detection with Photonic QCNNs:
Quantum Machine Learning (QML) is widely regarded as one of the most promising near-term applications of quantum computing. In this work, we introduce a photonic quantum convolutional neural network (QCNN) architecture designed for classification tasks. Our approach leverages the natural strengths of photonic hardware—including scalability, low energy consumption, and native implementation of linear transformations—to build a noise-resilient QML model. Benchmarking against classical baselines, we find that photonic QCNNs can achieve competitive accuracy. Crucially this benchmark shows that when complexifying the dataset, we have an advanatage for quantum models in the number of needed parameters to reach a given accuracy. To explore concrete applications, we are partnering with Orange to investigate the use of photonic QCNNs for malware detection in cybersecurity. Malware search and classification represent a critical industrial challenge, where fast and energy-efficient pattern recognition can deliver substantial business value. By combining photonic hardware with advanced QML architectures, our collaboration aims to demonstrate how quantum computing can address real-world security threats while maintaining sustainability and scalability.
q2b.qcware.com
