Semantic segmentation in remote sensing is commonly addressed using classical deep learning architectures such as U-Net, which require a large number of parameters to model complex spatial relationships. Quantum machine learning (QML) provides an alternative representation paradigm by mapping classical features into quantum states, but its direct application to high-dimensional images remains challenging under near-term quantum hardware constraints. In this work, we propose HQ-UNet, a hybrid quantum-classical U-Net architecture that integrates a compact parameterized quantum circuit at the bottleneck of a classical U-Net. The proposed design uses a non-pooling quantum convolutional module to enrich highly compressed encoder features before decoding, while keeping the quantum component shallow and parameter-efficient. Experiments on the LandCover.ai dataset show that HQ-UNet achieves a mean IoU of 0.8050 and an overall accuracy of 94.76%, outperforming the classical U-Net baseline. These results suggest that compact quantum bottlenecks can enhance feature representation for remote sensing image segmentation under near-term quantum constraints. This highlights the potential of hybrid quantum-classical designs as a promising direction for parameter-efficient dense prediction in Earth observation.
Semantic segmentation in remote sensing is commonly addressed using classical deep learning architectures such as U-Net, which require a large number of parameters to model complex spatial relationships. Quantum machine learning (QML) provides an alternative representation paradigm by mapping classical features into quantum states, but its direct application to high-dimensional images remains challenging under near-term quantum hardware constraints. In this work, we propose HQ-UNet, a hybrid quantum-classical U-Net architecture that integrates a compact parameterized quantum circuit at the bottleneck of a classical U-Net. The proposed design uses a non-pooling quantum convolutional module to enrich highly compressed encoder features before decoding, while keeping the quantum component shallow and parameter-efficient. Experiments on the LandCover.ai dataset show that HQ-UNet achieves a mean IoU of 0.8050 and an overall accuracy of 94.76%, outperforming the classical U-Net baseline. These results suggest that compact quantum bottlenecks can enhance feature representation for remote sensing image segmentation under near-term quantum constraints. This highlights the potential of hybrid quantum-classical designs as a promising direction for parameter-efficient dense prediction in Earth observation.
Magnetic resonance imaging (MRI) brain tumor segmentation is critical for diagnosis and treatment planning, but remains challenging due to blurry boundaries and class imbalance. Although deep learning models like U-Net have set benchmarks, they often face limitations in modeling complex feature interactions and require substantial computational resources. This paper introduces Quantum U-Net, a novel hybrid quantum-classical architecture that augments the classical U-Net by replacing its final convolutional layer with a tailored quantum convolutional layer. This layer utilizes amplitude encoding to map 64-channel feature maps onto the state of 6 qubits, followed by a parameterized strongly entangling layer to exploit quantum superposition and entanglement for enhanced multi-channel feature fusion. We employ a two-stage training strategy: pre-training a classical U-Net for 50 epochs, freezing its parameters, and then fine-tuning the integrated quantum layer for 30 epochs. Evaluated on the Brain Tumor Segmentation dataset, our hybrid model achieves a Dice coefficient of 0.806, a significant 6.2% relative improvement over the baseline classical U-Net (Dice: 0.759). This work validates the potential of quantum-enhanced models to improve performance in medical image segmentation and provides a practical blueprint for hybrid quantum-classical architectures in resource-intensive domains.
Image segmentation represents a fundamental computer vision task, yet classical Fully Convolutional Networks (FCNs) often struggle to capture long-range dependencies and global context due to the limited receptive fields of local convolution operations. While incorporating fully connected layers at the bottleneck can address this limitation by enabling global information mixing, it introduces a prohibitive quadratic parameter growth scaling with the spatial resolution of the bottleneck feature map. We propose a hybrid quantum-classical UNet architecture that replaces the bottleneck layer with an eight-qubit variational quantum circuit, achieving constant parameter count regardless of input resolution while maintaining competitive segmentation performance. Through comprehensive experiments on synthetic segmentation tasks and validation on real-world ISIC 2016 medical lesion images, we demonstrate that quantum bottleneck layers achieve intersection-over-union of 0.8544 on synthetic data and 0.8644 on authentic medical imaging with expert annotations, representing parameter compression ratio of sixty-four fold compared to classical dense bottlenecks at 512×512 resolution. Theoretical kernel execution time analysis based on documented gate times suggests potential acceleration at standard resolutions, with asymptotic advantage growing as quantum hardware matures and overhead diminishes. Our theoretical analysis establishes that quantum circuits achieve logarithmic parameter scaling versus classical quadratic scaling for dense global context layers, with compression advantages growing exponentially as image resolution increases. Thiswork demonstrates the theoretical foundations for quantum-enhanced image segmentation with empirical validation on authentic clinical data, opening pathways for efficient deployment in resource-constrained medical imaging environments and high-resolution applications as quantum hardware infrastructure evolves.
Acceleration of image segmentation using deep learning methods on satellite imagery has become ubiquitous in various applications areas such as land cover classification, disaster monitoring and vegetation detection. However, the increase in satellite image resolution and large data volume required for remote sensing applications has resulted in a substantial increase in computational resource usage and demand for real-time processing. This paper investigates quantum computing as a novel approach to meet these computational demands, exploiting its parallel processing strengths. We evaluate hybrid quantum models (COQCNN, MQCNN, FQCNN) against classical CNN and U-Net architectures in remote sensing classification. Although COQCNN and MQCNN underperformed, FQCNN reached 53.26% accuracy, outperforming the classical CNN by 8%. Despite quicker convergence, quantum models struggle with complex feature segmentation, a task where U-Net excels. This study highlights quantum convolutions as a potential path to enhance convergence while addressing challenges like noise from multiple quantum channels affecting accuracy.
Abstract We introduce quantum feature extraction (QuFeX), a novel quantum machine learning module. The proposed module enables feature extraction in a reduced-dimensional space, significantly decreasing the number of parallel evaluations required in typical quantum convolutional neural network (NNs) architectures. Its design allows seamless integration into deep classical NNs, making it particularly suitable for hybrid quantum–classical models. As an application of QuFeX, we propose Qu-Net—a hybrid architecture which integrates QuFeX at the bottleneck of a U-Net architecture. The latter is widely used for image segmentation tasks such as medical imaging and autonomous driving. Our numerical analysis indicates that the Qu-Net can achieve superior segmentation performance compared to a U-Net baseline. These results highlight the potential of QuFeX to enhance deep NNs by leveraging hybrid computational paradigms, providing a path towards a robust framework for real-world applications requiring precise feature extraction.
The characterization of brain tumors from magnetic resonance imaging (MRI) is crucial for diagnosis, treatment planning, and prognosis. But the analysis of brain tumors based on MRI is still difficult because of the significant intra-tumoral heterogeneity, ambiguous boundaries of lesions, high dimensionality of imaging features, and complicated nonlinear relationships between anatomical structures. While transformer-based deep learning (DL) models have greatly advanced automated tumor analysis, classical feature representations often fail to effectively capture complex feature interactions in pathological images, especially when distinguishing between subtle pathological variations and normal tissue needs to be done efficiently. Recent quantum machine learning (QML) developments indicate that quantum feature encoding can map complex data to exponentially larger Hilbert spaces, leading to a more expressive representation of nonlinear patterns and feature correlations that are challenging to model with conventional architectures. Encouraged by this potential, this study proposes a hybrid quantum-classical framework, named QFormer-Brain, which combines transformer-based segmentation with quantum representation learning for automated brain tumor segmentation and classification. To segment the tumor regions, a Shifted Window UNet Transformers (Swin-UNETR) architecture is first used to capture both local anatomical details and global contextual information through hierarchical self-attention. Then, the deep semantic features of the segmented lesions are mapped to the quantum feature mapping module with angle encoding and variational quantum circuits (VQCs), which can obtain richer nonlinear feature representation by quantum superposition and entanglement mechanisms. The obtained quantum embeddings are then combined with features obtained from the transformer and classified by a Quantum Transformer Classifier. The experiments were performed on the publicly available BraTS 2021 data set using an 8-qubit variational quantum circuit, with 1024 measurement shots and readout-error mitigation, via the IBM Qiskit Aer simulator. The proposed framework achieved a Dice Similarity Coefficient (DSC) of 97.1%, Intersection-over-Union (IoU) of 95.0%, classification accuracy of 98.5%, sensitivity of 98.1%, specificity of 99.0%, and F1-score of 98.3%, outperforming U-Net, nnU-Net, ResNet50, Vision Transformer (ViT), and conventional hybrid Convolutional Neural Network (CNN)-based models. These results show that quantum-inspired representation learning can be efficiently integrated into transformer-based models to boost the discriminative feature extraction from complex MRI data.
Abstract The U-Net architecture is widely recognized as one of the most prominent models for medicalimage segmentation, comprising an encoder and a decoder. The encoder is crucial for extractingimage features to enhance segmentation accuracy, typically incorporating convolutional and poolinglayers. However, standard encoder structures often miss some classical features and struggleto extract high-level features effectively. In this paper, a hybird Quantum-Classical Neural Network(QC-Net) model with a novel encoder is proposed, aiming to capture more representativefeatures. Our encoder features a Residual Convolutional Block (RCB) to primarily extract somemissed features, and then Efficient Channel Attention (ECA) is employed into the output featuremaps after the RCB and pooling operations to handle more complex and noisy information. Consequently,a two-qubit parameterized circuit is devised to capture the final output features of theencoder, aiming to further capture the hidden high-level features from the quantum dimension.The decoder incorporates a joint attention mechanism and deconvolution operations to recoverthe spatial resolution and detail information of the original input image. To validate its efficacy,we conduct skin lesion segmentation experiments utilizing the ISIC2018 dataset. Notably, ourQC-Net model outperforms both U-Net and CA-Net, achieving an average Dice coefficient of93.2%, indicating improvements of 5.43% and 1.12%, respectively. These results underscore theoutstanding performance of our proposed QC-Net model.