.png){kind=icon}
.png){kind=icon}
Return
Hybrid quantum-classical 3D object detection using multi-channel quantum convolutional neural network
.png){kind=icon}
PRE
AI
DOI：10.1007/s11227-025-06968-7.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
3D object detection has recently shown remarkable progress in the computer vision field, enabling advanced understanding of the surrounding environment by identifying objects' shape, position, and depth. However, processing high-dimensional data using classical convolutional neural networks (CNNs) introduces considerable computational challenges. This paper proposes a novel hybrid quantum-classical 3D object detection (HQCOD) approach, integrating a multi-channel quantum convolutional neural network (MC-QCNN) to significantly reduce computational complexity by leveraging quantum computing advantages. Additionally, knowledge distillation (KD) is applied to enhance detection accuracy and model robustness. Experimental evaluations using the Karlsruhe institute of technology and Toyota technological institute (KITTI) dataset validate the scalability and effectiveness, demonstrating the HQCOD as a practical quantum-assisted solution. This study establishes a foundation for extending quantum-enhanced 3D computer vision methods into real-world applications.
Keywords:
Quantum machine learning
Multi-channel quantum convolutional neural network
Hybrid quantum-classical 3D object detection
Quantum keypoint feature pyramid network
Knowledge distillation