WiMi Hologram Cloud Inc. (NASDAQ: WiMi) is developing a data compression technology that integrates the Quantum Haar Transform (QHT) with quantum partial measurement technology to create a novel approach. The company reports this approach constructs correlations between feature dimensions through quantum entanglement, preserving data’s structural information while compressing it. Built on the universal quantum circuit of the quantum Fourier transform, QHT uses superposition to efficiently transform high-dimensional data, aiming to solve the computational complexity issues faced by classical methods in high-dimensional data processing.
Quantum Haar Transform Enables High-Dimensional Data Mapping
WiMi Hologram Cloud Inc. utilizes the Quantum Haar Transform (QHT) to map high-dimensional classical data to a quantum state space, with each qubit representing a feature dimension and superposition coefficients encoding feature intensity. This approach addresses the exponential increase in computational complexity faced by classical Haar transforms when processing high-dimensional data. Quantum partial measurement technology complements the QHT by selectively extracting key feature information from the quantum state, based on the probabilistic interpretation of quantum states, rather than discarding data as in classical pooling strategies.
If a max-pooling strategy is adopted, the measurement basis construction aims to maximize the collapse probability of the quantum state corresponding to the maximum feature intensity; conversely, average-pooling achieves a weighted average of feature intensities through measurement basis orthogonality. Unmeasured qubits remain in superposition, maintaining local feature correlations, while measurement results output low-dimensional classical feature vectors, optimizing feature preservation and dimension compression.
The technology’s core relies on Variational Quantum Algorithms (VQA), a hybrid quantum-classical approach consisting of a Parameterized Quantum Circuit (PQC) and a classical optimizer. The classical optimizer iteratively adjusts PQC parameters to minimize a loss function, such as feature reconstruction error, enabling precise control of quantum state transformation.
This allows optimization of QHT quantum gate parameters to maximize local feature correlation preservation when mapping high-dimensional data and to adapt pooling output features to downstream tasks. This system exhibits efficient computational characteristics; it uses quantum parallelism and QHT orthogonality to achieve polynomial-level reduction in complexity compared to classical high-dimensional data pooling algorithms, significantly improving processing efficiency for large-scale data.
See today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals.
