The emergence and advancement of new technologies present promising opportunities for data-driven research in the Digital Humanities (DH). As an innovative intersection of quantum computing and humanities data, Quantum Humanities (QH) holds significant research potential. However, current studies related to QH remain scarce with limited applications involving quantum deep learning, and its feasibility, effectiveness, and efficiency on humanities data need in-depth exploration. To address this gap, this paper takes the role type recognition of painted faces of Beijing Opera as a case study. Several classical and competitive quantum deep learning–based and classical deep learning–based models (e.g., quantum convolutional neural networks and classical convolutional neural networks) are selected for comparison on the public image dataset Painted Faces of Beijing Opera digitized from books. Extensive empirical experimental results demonstrate that quantum deep learning–based models are of certain advantages and hold promising prospects in applied DH practices.
The emergence and advancement of new technologies present promising opportunities for data-driven research in the Digital Humanities (DH). As an innovative intersection of quantum computing and humanities data, Quantum Humanities (QH) holds significant research potential. However, current studies related to QH remain scarce with limited applications involving quantum deep learning, and its feasibility, effectiveness, and efficiency on humanities data need in-depth exploration. To address this gap, this paper takes the role type recognition of painted faces of Beijing Opera as a case study. Several classical and competitive quantum deep learning–based and classical deep learning–based models (e.g., quantum convolutional neural networks and classical convolutional neural networks) are selected for comparison on the public image dataset Painted Faces of Beijing Opera digitized from books. Extensive empirical experimental results demonstrate that quantum deep learning–based models are of certain advantages and hold promising prospects in applied DH practices.
Abstract: Deep learning algorithms have shown promising results for different image processing tasks, particularly in remote sensing & image recognition. Till now many studies have been carried out on image processing, which brings a new paradigm of innovative capabilities under the umbrella of intelligent remote sensing and computer vision. Accordingly, quantum processing algorithms have proved to efficiently solve some issues that are undetectable to classical algorithms and processors. Keeping that in mind, a Quantum Convolutional Neural Network (QCNN) architecture along with Hybrid Quantum filters would be utilized supported by cloud computing infrastructures and data centers to provide a broad range of complex AI services and high data availability. This research summaries the conventional techniques of Classical and Quantum Deep Learning and it’s research progress on realworld problems in remote sensing image processing as a comparative demonstration. Last but not least, we evaluate our system by training on Street View House Numbers datasets in order to highlight the feasibility and effectiveness of using Quantum Deep Learning approach in image recognition and other similar applications. Upcoming challenges and future research areas on this spectrum are also discussed.
The rapid advancements in quantum computing have opened new avenues for enhancing classical machine learning paradigms, particularly in the realm of image classification. Traditional Deep Convolutional Neural Networks (DCNNs) have achieved remarkable success, yet they face challenges related to computational intensity and the need for vast datasets [4, 19, 31]. Quanvolutional Neural Networks (QNNs) emerge as a promising hybrid quantum-classical approach that integrates quantum circuits directly into the feature extraction process of convolutional layers. This article explores the fundamental characteristics and operational advantages of QNNs, focusing on how their unique quantum-enhanced feature maps contribute to improved image classification performance. We delve into the architecture of quanvolutional layers, the mechanisms of data encoding, and the potential for quantum advantage in feature learning. By synthesizing recent research, we demonstrate the theoretical underpinnings and observed benefits of QNNs in extracting richer, more discriminative features, potentially leading to higher accuracy and efficiency, especially in the Noisy Intermediate-Scale Quantum (NISQ) era [22]. Challenges such as data encoding complexity, parameter optimization, and hardware limitations are also discussed, alongside future directions for scalable and robust QNN implementations.
In the global field of computational exploration, quantum computing stands as an enigmatic force, poised to transcend classical limitations in certain tasks. Recent years have borne witness to an unprecedented surge in quantum computing research, fostering a vibrant ecosystem of inquiry and innovation. This study delves into the complex fabric of quantum computing, revealing its evolution from theoretical inception to blooming practical applications. It defines the evolution of quantum computing and paves a roadmap for its potential future through a detailed examination of cutting-edge research initiatives. This study takes readers on a tour through the history of quantum computing, revealing the rich fabric of technological advances and theoretical achievements. From the early stages of quantum hardware creation to the refinement of quantum algorithms and error mitigation measures, the story unfolds with utmost clarity. Furthermore, it deciphers the mysterious mechanics of found quantum algorithms, contrasting their abilities with regular computational paradigms. This paper aims to cast light on a largely untrod territory of interdisciplinary cooperation: the intersection of quantum computing and deep learning. It paves the way to a new generation of computational advantage through the incorporation of quantum rules and deep learning platforms where its application can redefine model enhancement, algorithms formulation and data modeling. Concepts such as quantum neural networks and quantum-enhanced optimization, paint the picture of potentials yet to be imagined inside this mutually beneficial relationship. Moving from conjecture consideration, this paper aims to discuss how quantum computing supports deep learning concretely in numerous application areas. Deep learning utilizing quantum computing has the possibility of transcending the previous boundaries in such areas as image identification, drug design and additional optimization. This investigation maps out the way forward for pragmatic realization by deconstructing existing applications and speculatively constructing new ones. At the end of the voyage, the paper looks forward and presents a roadmap to the next phase of voyaging and inventing. The requirement for a cross-disciplinary approach is then presented, and quantum computing scientists and deep learners are encouraged to collaborate. Through this process of collaboration and creativity the unison of quantum computers with deep learning represents the pursuit of further boundaryless violations as the next step to the advancement of tremendously significant scientific accomplishment in the future.
The rapid growth of computer vision and increasingly complex image recognition tasks has exposed fundamental computational limitations of classical machine learning models, motivating the exploration of quantum computing as an emerging new paradigm. This paper presents a comprehensive benchmarking study of classical and quantum machine learning models for image recognition on the MNIST handwritten digit dataset, evaluating both traditional models, a Classical Support Vector Machine (CSVM) and a Quantum Support Vector Machine (QSVM), and deep neural network models, a Classical Convolutional Neural Network (CCNN) and a Quantum Convolutional Neural Network (QCNN), across four performance dimensions: classification accuracy, computational runtime, parameter count, and memory requirements. Experiments are conducted as functions of both feature dimensionality and sample size, and across CPU and GPU execution environments, providing a controlled, multidimensional comparison to address gaps in prior work. For the SVM-based models, QSVM consistently outperforms CSVM in accuracy, reaching ∼ 0.90 versus ∼ 0.85 at 1,000 samples, with a higher computational cost. A feature count of 10 qubits and a sample size in the range of 200 -- 500 emerge as practical operating points that balance accuracy and runtime. For the neural network models, CCNN and QCNN achieve comparable classification accuracy, both exceeding 0.96 at 64 features and 60,000 samples, yet QCNN offers superior parameter and memory efficiency at higher feature counts, while incurring higher runtime. Across both model families, quantum models consistently outperform classical models by greater margins in accuracy as feature dimensionality or sample size increases.
The integration of quantum computing principles with classical deep learning architectures has emerged as a promising direction for advancing image recognition systems beyond the limitations of purely classical convolutional neural networks (CNNs). This study proposes and analyzes a hybrid quantum–classical deep learning framework based on quanvolutional neural network (QNN) principles, where quantum circuits are embedded within convolutional feature extraction layers to enhance representation capability. The research investigates how quantum-enhanced filters, particularly quanvolutional layers, contribute to improved feature diversity, non-linear transformation capacity, and classification accuracy in image recognition tasks. Building upon established convolutional architectures and quantum machine learning theories, the proposed framework integrates quantum random circuit encoding with classical deep neural networks to evaluate performance gains in comparison to conventional CNN models. The study further incorporates theoretical insights from hybrid quantum-classical learning systems, emphasizing the role of parameterized quantum circuits in feature mapping and optimization stability. The methodology is grounded in NISQ-era quantum computing constraints and leverages hybrid optimization strategies to ensure computational feasibility. Experimental evaluation demonstrates that quanvolution-based models can provide improved robustness in feature extraction, particularly in high-dimensional image spaces, while maintaining competitive computational efficiency. However, scalability limitations and noise sensitivity in quantum circuits remain key challenges. Overall, this work contributes to the growing field of quantum-enhanced machine learning by providing a structured performance analysis of hybrid quanvolutional architectures and highlighting their potential role in next-generation intelligent visual recognition systems.
Nowadays, with the discovery of the power and potential of quantum computers, developing and understanding quantum-based deep learning models has become an important research area. This study investigates Quantum Transfer Learning and Quantum Hybrid Learning models that involve feature extraction and training processes using Convolutional Neural Networks (CNN) and Vision Transformer (ViT). The study aims to explore the potential advantages and differences of quantum deep learning techniques. It is envisioned that quantum computing can provide significant advantages in terms of computational speed and efficiency, especially in complex and large-scale data sets. Therefore, this study will contribute to a better understanding of the practical applications and potential impacts of quantum deep learning techniques. In this study, we evaluate the performance of four different quantum deep learning architectures using two different datasets. The classifiers used are the pre-trained ResNet-50 with a kernel size of 5x5 and the state-of-the-art CaiT-24-XXS-224 (CaiT) transducers. Optimization was performed with Adam optimizer using the cross entropy loss function. A total of eight models were trained, each with ten iterations. Accuracy (Acc), balanced accuracy (BA), overall F𝛽 (F_beta) macro score F1 and F2, Matthew's Correlation Coefficient (MCC), sensitivity (Sens) and specificity (Spec) were used as performance measures.