Beijing, Feb. 06, 2026 (GLOBE NEWSWIRE) -- WiMi Releases Hybrid Quantum-Classical Neural Network (H-QNN) Technology for Efficient MNIST Binary Image Classification BEIJING, Feb.06, 2026––WiMi Hologram Cloud Inc. (NASDAQ: WiMi) ("WiMi" or the "Company"), a leading global Hologram Augmented Reality ("AR") Technology provider, today announced the release of a Hybrid Quantum-Classical Neural Network (Hybrid Quantum-Classical Neural Network, H-QNN) technology for efficient MNIST binary image classification. This ...

Core Idea - WiMi Hologram Cloud Inc. has introduced a new Lean Classical-Quantum Hybrid Neural Network (LCQHNN) framework designed to enhance learning efficiency while minimizing quantum circuit complexity, marking a significant advancement in the practical application of quantum neural networks [1][11]. Technical Overview - The LCQHNN framework focuses on quantum feature amplification combined with classical stability optimization, creating an efficient interaction mechanism between classical and quantum computing [2]. - The architecture consists of a Classical Front-End for feature extraction and a Quantum Back-End utilizing variational quantum circuits for classification [2]. - The classical component employs lightweight convolutional layers for data preprocessing, embedding results into quantum state space for feature transformation [3]. - The quantum section features a four-layer variational quantum circuit (4-layer VQC) that optimizes circuit parameters to enhance classification performance while reducing resource consumption [4]. Workflow Stages - The workflow includes several key stages: 1. Data Preprocessing and Classical Encoding, where images are processed into medium-dimensional vectors for quantum encoding [5]. 2. Quantum State Preparation and Entanglement Structure Construction, enhancing correlations between qubits to improve model performance [6]. 3. Parameterized Quantum Evolution and Measurable Readout, utilizing adjustable parameters for efficient training and measurement [7]. 4. Classical Feedback and Hybrid Optimization, coordinating classical and quantum parameter updates to minimize classification errors [8]. 5. Classification Decision and Feature Visualization, where final results are decoded back to the classical domain, demonstrating strong inter-class separability [9]. Future Directions - WiMi plans to expand the LCQHNN model into multimodal learning, integrate with quantum support vector machines and quantum convolutional networks, and promote prototype deployment on quantum hardware to validate performance in real-world scenarios [10]. - The company aims to combine quantum parallel optimization with federated learning frameworks to develop secure and efficient quantum intelligent systems [10]. Company Background - WiMi Hologram Cloud Inc. specializes in holographic cloud services, focusing on various professional applications including AR technologies, holographic devices, and metaverse solutions [12].

Company Overview - WiMi Hologram Cloud Inc. is a leading global provider of Hologram Augmented Reality (AR) technology, focusing on various professional fields including in-vehicle AR holographic HUD, 3D holographic pulse LiDAR, and metaverse holographic AR/VR devices [1][15]. Technology Development - WiMi has proposed a new hybrid quantum-classical learning technology based on quantum convolutional neural networks (QCNN), which significantly improves performance in multi-class image classification tasks by recycling discarded qubit state information [2][3]. - This technology optimizes the efficiency of quantum networks under noisy intermediate-scale quantum (NISQ) devices and demonstrates the potential for quantum information reuse, paving a new path for hybrid quantum-classical models [3][12]. Quantum Computing Insights - Traditional QCNNs face challenges such as increased training time and energy consumption as model depth increases, necessitating a rethink of intelligent computing architectures [4][5]. - WiMi's architecture allows for the re-participation of discarded qubits in decision-making, enhancing the model's overall expressive ability and compensating for information loss during the pooling stage [6][10]. Architectural Innovation - The proposed architecture features a dual-channel structure that utilizes both retained and discarded qubit information, achieving maximum utilization of quantum information at the feature level [7][9]. - The training process employs a joint optimization mechanism based on classical cross-entropy loss, integrating outputs from both quantum and classical layers for improved performance [11][12]. Future Implications - WiMi's hybrid quantum-classical learning technology represents a significant advancement in quantum intelligence, exploring practical paths under current NISQ constraints and demonstrating the potential of quantum machine learning in various applications [13][14]. - As quantum computing progresses towards practical applications, hybrid models are expected to bridge the gap between theory and industry, driving innovation in fields such as intelligent vision and medical diagnosis [14].

Core Viewpoint - MicroCloud Hologram Inc. has launched a multi-class classification method based on Quantum Convolutional Neural Network (QCNN) using hybrid quantum-classical learning, showcasing the potential of quantum computing in artificial intelligence and addressing limitations of traditional computing architectures [1][12]. Technology Development - The new technology addresses bottlenecks in classical neural networks, particularly in computing power, energy consumption, and model complexity, as data scales and classification categories expand [2][10]. - The core of the technology is a multi-class classification model that integrates quantum convolutional neural networks with a hybrid quantum-classical optimization framework, utilizing the TensorFlow Quantum platform [3][9]. Model Design - HOLO introduced a quantum perceptron model that leverages quantum state evolution and measurement, enhancing feature extraction through quantum gates, which allows for high-dimensional feature mappings [4][9]. - The model optimizes circuit complexity and improves expressive power by reducing redundant gate operations and enhancing entanglement structures [4][9]. Training Mechanism - The hybrid quantum-classical learning mechanism significantly improves training efficiency and model convergence speed by combining quantum state encoding with classical optimization algorithms [5][12]. - The training process involves quantum circuits encoding input samples, producing measurement results that are normalized and used to update quantum circuit parameters iteratively [5][8]. Experimental Results - Experimental results indicate that HOLO's quantum convolutional neural network achieves accuracy comparable to classical convolutional networks in four-class classification tasks, validating the feasibility of quantum neural networks for practical applications [6][12]. Industry Context - The technology is positioned to address challenges in various fields such as computer vision, medical image analysis, and natural language processing, where traditional deep learning methods face limitations due to high energy consumption and long training times [10][12]. - HOLO's quantum convolutional neural network method aims to reduce computational complexity and leverage future advancements in quantum hardware for significant breakthroughs in computing power [10][11]. Future Prospects - The technology lays the groundwork for broader applications of quantum machine learning, with expectations for expansion into large-scale image recognition and real-time video processing as quantum hardware advances [11][12]. - HOLO plans to further optimize quantum circuits and explore multi-layer quantum convolutional networks combined with deep residual structures in future research [11].

Core Viewpoint - MicroCloud Hologram Inc. has introduced a groundbreaking technology, the Multi-Class Quantum Convolutional Neural Network (QCNN), which aims to enhance multi-class classification tasks by leveraging the advantages of quantum computing, marking a significant step towards the practical industrialization of quantum machine learning [1][9]. Technology Development - The QCNN technology integrates quantum algorithms with convolutional neural networks, achieving superior performance in complex classification tasks compared to traditional neural networks, especially as the number of categories increases [1][2]. - QCNN utilizes parameterized quantum circuits to simulate core operations of convolutional neural networks, enabling efficient modeling of complex feature distributions through quantum gate operations and entangled states of qubits [3][4]. Technical Implementation - The training of QCNN differs from classical neural networks by optimizing parameterized quantum circuits using a cross-entropy loss function and the PennyLane framework for automatic differentiation [4]. - Two optimization methods are employed: one based on polynomial approximations for high-precision gradient estimation and another using finite difference methods for computational flexibility, enhancing training convergence and addressing gradient vanishing issues [4][5]. Computational Efficiency - QCNN demonstrates remarkable computational efficiency, particularly in scenarios with fewer parameters, leading to shorter training times and potential advantages in energy consumption and cost control as quantum hardware advances [5][6]. Strategic Importance - The development of QCNN is a strategic move for MicroCloud, positioning the company at the forefront of the quantum computing revolution, with applications anticipated in fields such as speech recognition, medical diagnostics, financial risk control, and autonomous driving [6][7]. - The technology is seen as a cornerstone of the company's quantum intelligence strategy, aiming to build an intelligent computing platform that redefines the logic of intelligent computing beyond traditional artificial intelligence [8][9].

Core Insights - WiMi Hologram Cloud Inc. has announced a study on multidimensional pooling optimization techniques in variational quantum algorithms, introducing the Quantum Haar Transform (QHT) as a novel solution for multidimensional data pooling [1][4]. Group 1: Quantum Technology and Applications - The Quantum Haar Transform (QHT) maps multidimensional data to a quantum state space, enhancing the expression of local features while preserving the global structure of the data [2]. - Quantum partial measurement techniques allow for selective extraction of key information from quantum states, improving the pooling operation by retaining important feature information in a probabilistic form [2][3]. - Variational Quantum Algorithms (VQA) optimize parameters to ensure accurate pooling operations, maintaining computational efficiency and preserving locality and feature structure of the data [3][4]. Group 2: Advantages of Quantum Pooling Optimization - The multidimensional pooling optimization technology under the VQA framework addresses limitations of traditional pooling methods, leveraging quantum computing advantages for complex multidimensional data tasks [4]. - As quantum computing technology matures, this optimization technology is expected to demonstrate significant application potential across various fields, supporting the development of efficient quantum machine learning models [4]. Group 3: Company Overview - WiMi Hologram Cloud, Inc. specializes in holographic AR technologies, offering solutions in areas such as automotive HUD software, 3D holographic pulse LiDAR, and interactive holographic communication [5].

Company Overview - Spectral Capital Corporation (OTC: FCCN) is a leading developer and acquirer of AI and quantum technologies, with over 100 innovations in hybrid computing developed in 2025 alone, contributing to a broader portfolio of more than 500 patentable inventions [1][7] - The company specializes in the intersection of AI technology and quantum computing, leveraging over 20 years of expertise in accelerating emerging technologies [6][7] Innovations and Technologies - The focus of Spectral Capital's innovations is on Hybrid Quantum-Classical Algorithms, which are designed to improve computational efficiency and reduce operational costs for companies training large-scale AI models [1][2] - Hybrid Quantum-Classical Algorithms distribute workloads between classical and quantum computers, enhancing the performance of AI systems by refining parameters based on quantum outputs [3][4] - These innovations are expected to serve as a "force multiplier" across various AI-centric businesses, including telecommunications and fraud prevention, allowing for reduced cloud computing costs and increased learning speeds [5] Market Applications - The technology has broad applicability across various verticals, including messaging, predictive analytics, and intelligent infrastructure, where the company continues to acquire high-potential assets [2][5] - The collaborative computing architecture is central to emerging Quantum Machine Learning (QML) models, enabling more accurate predictions and accelerated learning cycles [4] Future Outlook - Spectral Capital is committed to expanding its innovation portfolio throughout 2025, focusing on commercializing key technologies in its active and planned acquisitions [5] - The company aims to leverage its deep patent portfolio to deliver compound value through technical differentiation and scalable platform deployment [5]

Core Insights - SEALSQ Corp is emphasizing the strategic advantages for Swiss financial institutions in adopting quantum technologies early, particularly through its collaboration with Wecan, a provider of secure digital infrastructure [1][6] - Quantum computing is positioned to transform the global financial system by overcoming limitations of traditional systems in risk modeling, option pricing, and fraud detection, offering exponential improvements in processing speed and analytical power [3][4] - The partnership between SEALSQ and Wecan aims to enhance cybersecurity and regulatory compliance through the integration of quantum-ready solutions, ensuring a secure transition into the quantum era [6][7] Industry Implications - Swiss banks and insurance firms are uniquely positioned to leverage quantum technologies due to their high standards of precision and security, which will be further strengthened in a post-quantum era [4][5] - Quantum machine learning techniques can address the evolving nature of financial fraud by uncovering anomalies in high-dimensional data spaces that traditional models struggle to detect [5] - The development of tokenized compliance frameworks, such as the Wecan Token, aims to streamline auditability and identity management while embedding quantum-resilient security standards [7][8] Company Developments - SEALSQ is pioneering Post-Quantum Technology solutions, focusing on quantum-resistant cryptography and semiconductors to address security challenges posed by advancing quantum computing [10][11] - The company's products are designed to protect sensitive data across various applications, enhancing resilience and security in industries such as healthcare, automotive, and defense [11] - SEALSQ's collaboration with Wecan is set to create a next-generation layer of digital trust for the financial industry, ensuring that institutions are prepared for future challenges [6][8]

Core Insights - MicroCloud Hologram Inc. has developed a noise-resistant Deep Quantum Neural Network (DQNN) architecture aimed at universal quantum computing and optimizing quantum learning tasks [1][13] - This architecture represents a significant advancement over traditional quantum neural networks by enabling efficient hierarchical training and reducing quantum resource demands [3][13] Group 1: Architecture and Functionality - The DQNN architecture utilizes qubits as neurons and arbitrary unitary operations as perceptrons, allowing for robust learning from noisy data [3][4] - Quantum neurons in this architecture are represented by quantum states, which can store richer information and enhance computational power through quantum superposition and entanglement [4] - Each neuron updates its state through unitary operations, preserving the normalization property of quantum states and ensuring information retention during computation [5] Group 2: Training and Optimization - HOLO employs an optimization strategy based on fidelity, aiming to maximize the similarity between current and target quantum states, which allows for fewer training steps and reduced quantum resource requirements [6][13] - This optimization method demonstrates strong robustness against noise and errors, maintaining stable learning performance even in noisy environments [7][11] Group 3: Scalability and Practical Applications - The architecture optimizes quantum state encoding, ensuring that the number of qubits scales with the network's width rather than its depth, thus reducing hardware demands [8][9] - HOLO's DQNN has shown excellent generalization capabilities, accurately learning target quantum operations and inferring reasonable quantum mapping relationships even with limited or noisy training data [10][11] - As quantum computing technology advances, the practical application prospects of deep quantum neural networks are broadening, with potential impacts across various industries [12][13]

Core Insights - The Unisys Innovation Program (UIP) recognized three teams for their innovative projects aimed at sustainable electric mobility, tumor diagnostics accuracy, and remote neurological monitoring [1][6]. Group 1: Unisys Innovation Program Overview - The UIP, established in 2009, connects engineering students with industry experts to develop technology-centric solutions [7]. - This year's program attracted over 890 project submissions from 12,760 registrants across eight themes addressing various business and technical challenges [2]. Group 2: Winning Projects - First place was awarded to RV College of Engineering for their project "Quantum Neural Network for Brain Tumor MRI Classification via Quantum Machine Learning," achieving over 98% accuracy in tumor diagnostics from lower-field (1.5T) scans, enabling cost-effective MRI analysis [3]. - Second place also went to RV College of Engineering for their project "Electronic Differential for Electric Vehicles," which integrates AI for real-time wheel speed control, enhancing traction, stability, and energy efficiency in electric vehicles [4]. - Third place was awarded to Ramaiah Institute of Technology for their "Earbud-Based Electroencephalogram (EEG) Monitoring System," a discreet wearable device for continuous brain monitoring using AI-based seizure detection [5]. Group 3: Leadership Insights - Unisys leadership emphasized the program's role in fostering creativity and technical excellence among India's engineering students, highlighting the importance of nurturing young talent [6][7].