In this industry-specific research topic, we foresee creating a system that would deliver a 360° perspective of non-verbal and behavioural aspects into insightful contexts by recording kinesics like facial expressions, emotions, gestures, and spatial attributes throughout interview/meeting room processes conducted digitally or in person.In this research, we investigate the transfer learning training accuracy and performance measures of various pre-trained Computer Vision model architectures with newly devised hybrid end-to-end Quantum Neural Network (QNN) and evaluate hybrid QNN model inference performance for Computer Vision applications deployed in the real world.
In this industry-specific research topic, we foresee creating a system that would deliver a 360° perspective of non-verbal and behavioural aspects into insightful contexts by recording kinesics like facial expressions, emotions, gestures, and spatial attributes throughout interview/meeting room processes conducted digitally or in person.In this research, we investigate the transfer learning training accuracy and performance measures of various pre-trained Computer Vision model architectures with newly devised hybrid end-to-end Quantum Neural Network (QNN) and evaluate hybrid QNN model inference performance for Computer Vision applications deployed in the real world.
Abstract Facial expressions bring human interactions to life with nonverbal cues that convey hues of emotions, feelings, and cultural intent. Analyzing these expressions is essential in the age of digital transformation. Traditional approaches to real-time facial expression analysis have limitations in capturing the complexity of these expressions and consume enormous computational power. The promising computing capacity of quantum computers is poised to solve current problems and set new standards by meeting growing demands. We have created a hybrid model that leverages the capabilities of the classical and quantum domains. The proposed model uses a quantum distance-based classifier along with classical artificial neural networks to perform real-time facial expression analysis and classification. Based on a thorough review of the literature, this novel and comprehensive work is, to the best of our knowledge, one of the few that addresses eight different emotions in the quantum domain. The implementation of a novel quantum error correction method has improved the accuracy of this hybrid model. Our model was trained on the CK+ facial expression database. To ensure fairness of the study, we tested our model on the largest facial expression database, AffectNet-8, and compared the performance with state-of-the-art models. Our model accuracy is 10.83% higher than that of the state-of-the-art models. From a holistic perspective, our proposed novel hybrid model appears to have universal value for all kinds of real-time image analysis and classification problems. As we move forward, we planned to focus on quantum neural networks for image processing and facial expression analysis.
Many security applications like face recognition and iris recognition developed to ensure the safety of the critical or personal data. Reliability of these applications are highly depending over the reliability of machine vision algorithms. Along with the development of the generative models standard machine vision dependent data security measurement systems became vulnerable. Currently, generating fake data to bypass machine vision dependent security systems is possible with a personal computer. In order to ensure the reliability of security measurements depending on the machine vision techniques it is critical to recognize fake images. In this research, possible use case of a quantum computer to ensure the reliability of machine vision dependent security systems is investigated. A hybrid quantum-classical hybrid model with the transfer learning approach is proposed to recognize whether if a face is fake or not. Effects of the quantum model’s depth over the accuracy is explored. ResNet-18 architecture is used as the classical part and a custom quantum neural network architecture built with the dressed quantum circuits is used as quantum part. This research is aimed to extend the use cases of quantum neural networks to security applications. Accuracies of quantum neural networks with different depths are reported. A simulated quantum computer is used to train the models. Along with the proposed approach it is concluded that it is possible to apply classical-quantum neural networks to improve the reliability of machine vision dependent security systems after the quantum co-processors become available in daily life.
The emerging classical-quantum transfer learning paradigm has brought a decent performance to quantum computational models in many tasks, such as computer vision, by enabling a combination of quantum models and classical pre-trained neural networks. However, using quantum computing with pre-trained models has yet to be explored in natural language processing (NLP). Due to the high linearity constraints of the underlying quantum computing infrastructures, existing Quantum NLP models are limited in performance on real tasks. We fill this gap by pre-training a sentence state with complex-valued BERT-like architecture, and adapting it to the classical-quantum transfer learning scheme for sentence classification. On quantum simulation experiments, the pre-trained representation can bring 50\% to 60\% increases to the capacity of end-to-end quantum models.
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.
Quantum computing, with its unique principles such as superposition and entanglement, promises to transform various domains, including artificial intelligence. This study investigates the fusion of quantum computing with contemporary AI technologies through quantum transfer learning. We introduce the Complex Quantum Kernel Self-Attention Network (CQKSAN), a novel model that leverages quantum circuits and self-attention mechanisms to boost text classification performance. By applying quantum transfer learning, CQKSAN harnesses the strengths of pre-trained classical models and enhances them with quantum capabilities. Experimental results on multiple binary classification datasets demonstrate that CQKSAN achieves a more efficient learning process. Specifically, it improves training accuracy and validation accuracy by up to 24.31% and 28.35%, respectively, and increases the Matthews Correlation Coefficient by 113.10%. Moreover, CQKSAN consistently outperforms the BERT baseline in both training and validation accuracy across various datasets, showcasing its strong learning capability and generalization performance in handling tasks with diverse linguistic structures and complex semantics. Our findings underscore the potential of hybrid quantum-classical approaches, while also identifying the current limitations and paving the way for future research in fully quantum-based learning systems.
Quantum computing is expected to fundamentally change computer systems in the future. Recently, a new research topic of quantum computing is the hybrid quantum–classical approach for machine learning, in which a parameterized quantum circuit, also called quantum neural network (QNN), is optimized by a classical computer. This hybrid approach can have the benefits of both quantum computing and classical machine learning methods. In this early stage, it is of crucial importance to understand the new characteristics of quantum neural networks for different machine learning tasks. In this paper, we will study quantum neural networks for the task of classifying images, which are high-dimensional spatial data. In contrast to previous evaluations of low-dimensional or scalar data, we will investigate the impacts of practical encoding types, circuit depth, bias term, and readout on classification performance on the popular MNIST image dataset. Various interesting findings on learning behaviors of different QNNs are obtained through experimental results. To the best of our knowledge, this is the first work that considers various QNN aspects for image data.