WiMi Hologram Cloud explores the technology of multivariate data pulling based on variable quantum algorithms. The proposed architecture combines Quantum Haar Transform, a partial quantum dimension and a hybrid scheme in which a parameterized quantum chain operates in conjunction with a classical optimizer. According to the company, this approach allows to reduce the dimension of high-size data while keeping local characteristics - without the need for a preliminary one-dimensional format.

In traditional machine learning, a pilling method is used to reduce the amount of data while maintaining key characteristics. WiMi states that a quantum diagram is capable of performing this operation directly for complex structures: audio-data, images, three-dimensional point clouds and hyperspectral arrays. However, the company did not specify a time frame for commercial implementation of the development.

WiMi Hologram Cloud is listed on the Nasdaq under the WIMI label and specializes in cloud holographic and AR solutions. Research in quantum computing complements the industry’s growing interest in hybrid algorithms: earlier this month, scientists from the Higher Technical School of Zurich presented a method for generating mathematically unimpeded randomness that addresses vulnerabilities in cryptography and digital systems.