WiMi Hologram Cloud Inc. launched a hybrid quantum neural network structure (H-QNN) for image multi-classification. This technology organically integrates the spatial feature extraction capabilities of classical convolutional neural networks (CNN) with the high-dimensional nonlinear mapping features of quantum neural networks (QNN), forming a new type of hybrid structure that possesses stronger generalization ability and computational efficiency in multi-class classification scenarios. This technology not only systematically optimizes the quantum-classical hybrid learning system in theory but also achieves classification accuracy and stability superior to similar algorithms in actual experiments, laying a solid technical foundation for quantum intelligent vision systems. The design of this hybrid quantum neural network (H-QNN) follows the principle of classical responsible for extraction and quantum responsible for discrimination.
The measurement probability distribution output by the quantum circuit is converted into feature vectors and fused with the output of the classical fully connected layer. This fused vector is input into the Softmax layer for final classification judgment. To further enhance the generalization performance in multi-class tasks, WiMi introduces a transfer learning mechanism, migrating the parameters of classical layers pre-trained in small-sample tasks to new tasks, thereby reducing the number of training epochs and enhancing model stability.
In actual implementation, this structure supports running on simulation environments and hardware quantum processing units (QPU). The simulation environment uses high-performance GPU clusters to complete training of classical modules, while quantum modules are executed in quantum simulators or FPGA-accelerated quantum kernel estimation environments, achieving heterogeneous collaboration of classical and quantum computing resources. By optimizing the cumulative variance contribution rate of PCA, it ensures that the mapping between input features and quantum state amplitudes maintains high information fidelity, thereby maximizing the utilization rate of quantum information.
Third, at the training strategy level, WiMi introduces a transfer Learning mechanism and parameter sharing structure. Traditional quantum neural networks often face risks of gradient disappear and overfitting in multi-class classification training, while parameter sharing can establish balanced gradient flow between different quantum layers, and the transfer learning mechanism enables the model to achieve rapid convergence on new tasks with fewer training epochs. In addition, WiMi designs an early stopping strategy based on the quantum Fidelity metric, which determines whether the training has reached the optimal point by monitoring the stability of quantum state evolution, thereby preventing overfitting.
Finally, at the system implementation level, it adopts a heterogeneous computing architecture, running the classical computing part on CPU/GPU platforms, while the quantum part is executed in quantum simulation modules implemented on FPGA. The FPGA module realizes reconfigurable execution logic for parameterized quantum circuits, capable of completing quantum state updates within nanosecond-level response times, thereby significantly improving the overall training speed of the system. This hybrid computing architecture demonstrates performance advantages far exceeding pure CPU or GPU simulations in experiments.
The proposal of WiMi's hybrid quantum neural network structure marks a key step in quantum artificial intelligence research moving from theoretical exploration toward practical applications. It not only demonstrates the potential advantages of quantum computing in the field of machine learning but also provides an engineered compromise solution for the current performance bottlenecks of quantum hardware. By embedding trainable quantum layers into the foundation of classical neural networks, this technology achieves efficient utilization of quantum computing resources, enabling quantum advantages to be embodied in real visual tasks.
In the future, quantum intelligence will no longer be merely a theoretical conception but will deeply integrate with fields such as deep learning, computer vision, and edge computing, becoming an important driving force for promoting the development of intelligent society. Let quantum intelligence move from the laboratory to the real world, and let quantum technology truly serve industrial upgrades and the expansion of human cognition.
