WiMi Develops Variational Quantum Algorithm-Driven Multi-Dimensional Data Pooling Optimization Technology

WiMi Hologram Cloud Inc. (NASDAQ: WiMi) announces a novel technology combining Variational Quantum Algorithms (VQA) with Quantum Haar Transform (QHT) and quantum partial measurement to optimize multi-dimensional data pooling. This approach preserves local feature correlations while achieving efficient dimension compression. The Quantum Haar Transform maps high-dimensional classical data into quantum states, encoding feature intensities via qubit superposition and entanglement, ensuring global structural preservation and local feature correlation reinforcement.

Quantum partial measurement then selectively extracts key feature information through probabilistic measurement bases, either maximizing feature intensity (max-pooling) or achieving weighted averages (average-pooling), maintaining local feature continuity. Variational Quantum Algorithms optimize quantum gate parameters for precise control of quantum state transformations, minimizing reconstruction errors or classification losses. This method avoids classical pooling’s information loss and significantly reduces computational complexity through quantum parallelism, enabling efficient processing of large-scale multi-dimensional data such as audio, images, point clouds, and hyperspectral data.

WiMi’s technology addresses the limitations of traditional pooling methods, unlocking quantum computing’s advantages for complex data tasks like computer vision, remote sensing, and biomedicine. The company’s focus on holographic cloud services includes in-vehicle AR HUDs, 3D holographic LiDAR, head-mounted light field devices, and metaverse holographic AR/VR solutions.

Source: Financial Times