QDS-SNN: Energy-efficient Quantum Deeply-Supervised Spiking Neural Network Algorithm for Traffic Sign Recognition

Researchers have proposed a new algorithm, QDS-SNN, for traffic sign recognition, achieving high accuracy while reducing energy consumption. The algorithm uses quantum superposition and entanglement to enhance performance.

The QDS-SNN algorithm, a Quantum Deeply-Supervised Spiking Neural Network, integrates Quantum Neural Networks (QNNs) for efficient deep supervision. It achieved 99.72% accuracy on the GTSRB dataset in 6 time steps, outperforming the MS-ResNet baseline by 1.32% while reducing energy consumption by 55.77%^[1]. On the TSRD dataset, QDS-SNN achieved 97.90% accuracy, reducing energy use to 52.68% of the baseline^[1]. In a related development, researchers have demonstrated an extension of hls4ml that enables the deployment of Spiking Neural Networks (SNNs) trained in PyTorch onto FPGA firmware. SNNs provide a naturally temporal machine-learning framework and are often associated with asynchronous neuromorphic processors. However, many real-time inference systems rely on conventional synchronous field-programmable gate arrays (FPGAs) and high-level synthesis (HLS) workflows. The authors demonstrated the workflow using a dense quantized SNN trained on the Heidelberg Spiking Digits dataset, achieving inference latencies of approximately 34μs^[2].