Similar Images Research Articles

Simultaneous multi-slice technique for reducing acquisition times in diffusion tensor imaging of the knee: a feasibility study.

To explore the feasibility of simultaneous multi-slice (SMS) technique for reducing acquisition times in readout-segmented echo planar imaging (RESOLVE) for diffusion tensor imaging (DTI) of the knee. A total of 30 healthy volunteers and 23 patients with knee acute injury (12 cases with anterior ligament (ACL) tears and 16 cases with patellar cartilage (PC) injury) were enrolled in this prospective study. Three DTI protocols were used: conventional RESOLVE-DTI with 12 directions (protocol 1), SMS-RESOLVE-DTI with 12 directions (protocol 2) and 20 directions (protocol 3). DTI parameters of gastrocnemius, ACL and posterior cruciate ligament (PCL), and PC from three protocols were quantitatively assessed. For volunteers, protocol 2 significantly reduced acquisition time by 38.6% and 34.2% compared to protocols 1 and 3 while maintaining similar high-quality images and similar diffusive parameters, except for the fractional anisotropy (FA) and axial diffusivity (AD) of the PC between protocols 2 and 1 (P < 0.05). For injured ACL and PC, protocols 1 and 2 showed similar accurate diffusive parameters (except for AD, P = 0.025) and similar diagnostic efficacy, which demonstrated significantly lower FA and higher radial diffusivity (RD) in protocols 1 and 2 compared to volunteers (P < 0.05). The 12-direction SMS-RESOLVE-DTI demonstrated a favorable balance between acquisition time and image quality, making it a promising alternative to conventional DTI for evaluating ligament and cartilage injuries. The SMS technique greatly reduces acquisition time while maintaining image quality, which signified the possibility of DTI's clinical application.

Deep learning-based automated detection and segmentation of bone and traumatic bone marrow lesions from MRI following an acute ACL tear

IntroductionTraumatic bone marrow lesions (BML) are frequently identified on knee MRI scans in patients following an acute full-thickness, complete ACL tear. BMLs coincide with regions of elevated localized bone loss, and studies suggest these may act as a precursor to the development of post-traumatic osteoarthritis. This study addresses the labour-intensive manual assessment of BMLs by using a 3D U-Net for automated identification and segmentation from MRI scans. MethodsA multi-task learning approach was used to segment both bone and BML from T2 fat‐suppressed (FS) fast spin echo (FSE) MRI sequences for BML assessment. Training and testing utilized datasets from individuals with complete ACL tears, employing a five-fold cross-validation approach and pre-processing involved image intensity normalization and data augmentation. A post-processing algorithm was developed to improve segmentation and remove outliers. Training and testing datasets were acquired from different studies with similar imaging protocol to assess the model's performance robustness across different populations and acquisition conditions. ResultsThe 3D U-Net model exhibited effectiveness in semantic segmentation, while post-processing enhanced segmentation accuracy and precision through morphological operations. The trained model with post-processing achieved a Dice similarity coefficient (DSC) of 0.75 ± 0.08 (mean ± std) and a precision of 0.87 ± 0.07 for BML segmentation on testing data. Additionally, the trained model with post-processing achieved a DSC of 0.93 ± 0.02 and a precision of 0.92 ± 0.02 for bone segmentation on testing data. This demonstrates the approach's high accuracy for capturing true positives and effectively minimizing false positives in the identification and segmentation of bone structures. ConclusionAutomated segmentation methods are a valuable tool for clinicians and researchers, streamlining the assessment of BMLs and allowing for longitudinal assessments. This study presents a model with promising clinical efficacy and provides a quantitative approach for bone-related pathology research and diagnostics.

High-quality AFM image acquisition of living cells by modified residual encoder-decoder network

Atomic force microscope enables ultra-precision imaging of living cells. However, atomic force microscope imaging is a complex and time-consuming process. The obtained images of living cells usually have low resolution and are easily influenced by noise leading to unsatisfactory imaging quality, obstructing the research and analysis based on cell images. Herein, an adaptive attention image reconstruction network based on residual encoder-decoder was proposed, through the combination of deep learning technology and atomic force microscope imaging supporting high-quality cell image acquisition. Compared with other learning-based methods, the proposed network showed higher peak signal-to-noise ratio, higher structural similarity and better image reconstruction performances. In addition, the cell images reconstructed by each method were used for cell recognition, and the cell images reconstructed by the proposed network had the highest cell recognition rate. The proposed network has brought insights into the atomic force microscope-based imaging of living cells and cell image reconstruction, which is of great significance in biological and medical research.

Kinship verification in childhood images using curvelet transformed features

Kinship verification poses a formidable challenge due to significant age disparities between parents and children, resulting in diminished facial resemblances and consequently reduced model accuracy. In response, we introduce a novel methodology focusing on childhood imagery, wherein images of both parents and children during their early stages are leveraged, leading to more pronounced feature similarities. Our approach introduces a novel feature descriptor utilizing the curvelet transform (CLT) to extract facial descriptors from sub-bands, effectively capturing unique facial features. Additionally, we employ a convolutional neural network (CNN) to extract texture information, further enhancing the discriminative power of our method. Statistical measures are subsequently applied to evaluate image similarity, with a dedicated training phase utilized to determine an optimal threshold for kinship classification during the testing phase. Extensive evaluations conducted across multiple databases, including KINFACEW-I, II, FIW, TSKINFACE, UBKinface, and childhood image dataset, showcase the superior performance of our proposed model compared to existing state-of-the-art approaches. Our methodology not only enhances kinship verification accuracy but also underscores the significance of childhood images in mitigating age-related challenges within the domain.

Machine Learning for Pedestrian-Level Wind Comfort Analysis

(1) Background: Artificial intelligence (AI) and machine learning (ML) techniques are being more widely employed in the field of wind engineering. Nevertheless, there is a scarcity of research on the comfort of pedestrians in terms of wind conditions with respect to building design, particularly in historic sites. (2) Objectives: This research aims to evaluate ML- and computational fluid dynamics (CFD)-based pedestrian wind comfort (PWC) analysis outputs using a novel method that relies on the sophisticated handling of image data. The goal is to propose a novel assessment method to enhance the efficiency of AI models over different urban scenarios. (3) Methodology: The stages include the analysis of climate data, CFD analysis with OpenFOAM, ML analysis using Autodesk Forma, and comparisons of the CFD and ML results using a novel image similarity assessment method based on the SSIM, MSE, and PSNR metrics. (4) Conclusions: This study effectively demonstrates the considerable potential of utilizing ML as a supplementary tool for evaluating PWC. It maintains a high degree of accuracy and precision, allowing for rapid and effective assessments. The methodology for precise comparison of two visual outputs in the absence of numerical data allows for more objective and pertinent comparisons, as it eliminates any potential distortions. (5) Recommendations: Additional research can explore the integration of ML models with climate data and different case studies, thus expanding the scope of wind comfort studies.

Integrated of Seismic Refraction Tomography (SRT), and Multichannel Analysis of Surface Waves (MASW) for Verifying and Mapping a postulated Fault

The location of the study site for the presented paper here is at Arkansas in the United States of America. Geophysical investigations using seismic refraction tomography (SRT) and multichannel analysis of surface waves (MASW) surveys have been jointly carried out on the basis of an experimental investigation in the field. The motive of this study is to validate and map a postulated fault, in the eastern part of Arkansas in the United State of America. The SRT techniques have shown their potentiality in successfully identifying the fault and generated remarkably similar images of the fault, while the MASW tool generated a slightly different image of the fault due to condition that the earth under survey made up of layers of material that increase in seismic velocity which consider the perfect case for SRT techniques. The results show that the integrated use of seismic refraction tomography and multichannel analysis of surface waves is an effective tool with great potential. They provide very useful data for shallow mapping applications.

Exploring the uncharted: adenoid cystic carcinoma nestled in temporal bone

BackgroundAdenoid cystic carcinoma is an uncommon malignancy primarily arising from salivary glands. An extremely rare site for adenoid cystic carcinoma is the skull base. We report a case of adenoid cystic carcinoma of skull base who presented with common complaints of pain and right ear discharge. The discussion is made with emphasis on imaging evaluation simulating infective etiology with adjacent skull base osteomyelitis. Careful observation of the imaging findings and further evaluation of the patient revealed the neoplastic nature of the lesion with the final diagnosis being adenoid cystic carcinoma.Case presentationA 40-year-old female presented to our department with complaints of pain and right ear discharge since 6 months with progressive, extensive facial swelling and facial nerve palsy. The patient had undergone modified radical mastoidectomy thrice, but the details were not available. On imaging, there was a heterogenous extensive lesion extending from scalp till upper cervical region with extensive destruction of skull base and intra-cranial extension. The possibilities of temporal bone squamous cell carcinoma and extensive skull base osteomyelitis were considered. Further the biopsy of the lesion revealed adenoid cystic carcinoma.ConclusionsExtensive lesions of the skull base can be of infective, neoplastic and inflammatory etiology. Distinguishing between these conditions is crucial, as they have similar imaging characteristics but require different management approaches. The presence of a lesion that displaces or destroys fascial planes, accompanied by solid mass-like enhancement, indicates a higher probability of a neoplastic origin rather than an infectious etiology. With squamous cell carcinoma being the most common neoplasm, adenoid cystic carcinoma of the skull base also needs to be understood due to its propensity for perineural spread and a high likelihood of recurrence.

Privacy-preserving geo-tagged image search in edge–cloud computing for IoT

The Internet of Things (IoT) generates a significant volume of geo-tagged images via surveillance sensors in edge–cloud computing environments. Image search is essential to facilitate information sharing, data analysis, and strategic decision-making. However, outsourced images are typically encrypted for privacy protection, posing a challenge in simultaneously searching for visual and geographical relevance on encrypted images. To address this, this paper proposes an edge intelligence empowered privacy-preserving top-k geo-tagged image search scheme for IoT in edge–cloud computing. The scheme presents a novel single-to-multi searchable encryption method for geo-tagged images that enables multiple users to perform secure nearest neighbor queries on a data source. Additionally, an extended anchor-based position determination method and an inner product-based distance calculation method are designed to enable geo-tagged image similarity calculation on ciphertext. Finally, a secure pruning method is introduced to improve query performance. Experiments are conducted to verify the performance of the scheme in terms of high efficiency and accuracy of the search.

Product Recommendation in Fashion E-Commerce using Deep Learning and Image Similarity

Product Recommendation in Fashion E-Commerce using Deep Learning and Image Similarity

Experience with a novel high frequency optical coherence tomography device for intracoronary imaging: a case series.

Intravascular imaging, especially optical coherence tomography (OCT), has significantly improved percutaneous coronary intervention (PCI), yet its routine clinical application faces challenges. This case series introduces the Gentuity® High-Frequency Optical Coherence Tomography (HF-OCT), a novel device designed to enhance intracoronary imaging with a significantly faster pullback and smaller catheter size, potentially offering enhanced navigability in complex lesions. We aimed to assess the image quality of Gentuity® HF-OCT in complex vessel conditions, as well as presenting a case series to illustrate the application of the device in various clinical scenarios. In this case series, we included all patients who underwent intracoronary HF-OCT imaging at our center. The primary endpoint was image quality assessed by clear image length (CIL). Image quality was assessed in relation to (1) lesion severity assessed by minimum lumen area (MLA); (2) vessel size, differentiating between larger (diameter ≥ 4 mm) and smaller vessel segments; (3) pre- vs. post-PCI conditions, and (4) vessel tortuosity, categorized into none, moderate, and severe. Twenty-four HF-OCT runs from 14 patients were included. No significant differences in CIL were observed across lesion severity terciles (p = 0.449), between small and large vessel segments [mean CIL% difference 1.3%; confidence interval (CI), -9.3 to 11.8; p = 0.802], and pre- vs. post-PCI conditions (mean CIL difference -3.9 mm; CI, -14.0 to 6.1; p = 0.373). Vessel tortuosity significantly impacted image quality, with clear reductions in CIL observed in cases of moderate (74.8; CI, 73.5 to 76.0; vs. 63.9; CI, 56.2 to 71.5; p = 0.043) and severe tortuosity (74.8; CI, 73.5 to 76.0; vs. 65.0; CI, 62.1 to 67.9; p = 0.002) compared to vessels with no tortuosity. Overall, the HF-OCT demonstrated excellent catheter deliverability and crossability, with very satisfactory image quality and no significant adverse events. The Gentuity® HF-OCT is a new OCT device capable of navigating both small- and large-diameter vessels, with similar image quality, but vessel tortuosity seems to have an impact on image quality. It appears to be as usable as conventional OCT for pre-PCI diagnosis and OCT-guided PCI, potentially bringing additional benefits in terms of deliverability, lesion crossover and ease of use in routine clinical practice.

Classification of Ameloblastoma, Periapical Cyst, and Chronic Suppurative Osteomyelitis with Semi-Supervised Learning: The WaveletFusion-ViT Model Approach.

Ameloblastoma (AM), periapical cyst (PC), and chronic suppurative osteomyelitis (CSO) are prevalent maxillofacial diseases with similar imaging characteristics but different treatments, thus making preoperative differential diagnosis crucial. Existing deep learning methods for diagnosis often require manual delineation in tagging the regions of interest (ROIs), which triggers some challenges in practical application. We propose a new model of Wavelet Extraction and Fusion Module with Vision Transformer (WaveletFusion-ViT) for automatic diagnosis using CBCT panoramic images. In this study, 539 samples containing healthy (n = 154), AM (n = 181), PC (n = 102), and CSO (n = 102) were acquired by CBCT for classification, with an additional 2000 healthy samples for pre-training the domain-adaptive network (DAN). The WaveletFusion-ViT model was initialized with pre-trained weights obtained from the DAN and further trained using semi-supervised learning (SSL) methods. After five-fold cross-validation, the model achieved average sensitivity, specificity, accuracy, and AUC scores of 79.60%, 94.48%, 91.47%, and 0.942, respectively. Remarkably, our method achieved 91.47% accuracy using less than 20% labeled samples, surpassing the fully supervised approach's accuracy of 89.05%. Despite these promising results, this study's limitations include a low number of CSO cases and a relatively lower accuracy for this condition, which should be addressed in future research. This research is regarded as an innovative approach as it deviates from the fully supervised learning paradigm typically employed in previous studies. The WaveletFusion-ViT model effectively combines SSL methods to effectively diagnose three types of CBCT panoramic images using only a small portion of labeled data.

SrSNet: Accurate segmentation of stroke lesions by a two-stage segmentation framework with asymmetry information

Ischemic stroke is a common neurological disease with a high mortality rate. Accurate stroke segmentation is crucial to evaluating patients’ disease outcomes. Stroke segmentation suffers from variable lesion size and noise interference due to similar image contrast characteristics of tissue structures. In this work, we propose a novel two-stage framework named Segmentation–reSegmentation Net (SrSNet) that mainly considers the bilateral symmetric comparison of brain hemispheres to locate pathological abnormalities. The whole framework includes two stages, the coarse segmentation stage and the fine segmentation stage. In the coarse segmentation stage, the brain images are first divided into three patches according to their spatial information to learn the local-level feature and then fed into a 2D Triplet Multi-scale Symmetric Transformer (TMSFormer) with a feature fusion module. This module facilitates the TMSFormer to generate detailed coarse segmentation results. In the fine segmentation stage, we refine the coarse segmentations with a 2D ResUNet. Extensive experiments are conducted using the two ischemic stroke lesion segmentation datasets, and the results are compared with other advanced models. Experiment results on two datasets show that SrSNet achieves state-of-the-art (SOTA) performance. Moreover, on the Ischemic Stroke Lesion Segmentation 2022 (ISLES’22) dataset, the recall score for stroke lesions that the maximum cross-sectional diameter is larger than 5 cm is 83.80%, and the recall score for strokes that the number of lesions of more than five is 79.05%. Overall, our method provides a potentially successful tool for improving the diagnosis of ischemic stroke.

Breast cancer detection through attention based feature integration model

&lt;span lang="EN-US"&gt;Breast cancer is detected by screening mammography wherein X-rays are used to produce images of the breast. Mammograms for screening can detect breast cancer early. This research focuses on the challenges of using multi-view mammography to diagnose breast cancer. By examining numerous perspectives of an image, an attention-based feature-integration mechanism (AFIM) model that concentrates on local abnormal areas associated with cancer and displays the essential features considered for evaluation, analyzing cross-view data. This is segmented into two views the bi-lateral attention module (BAM) module integrates the left and right activation maps for a similar projection is used to create a spatial attention map that highlights the impact of asymmetries. Here the module's focus is on data gathering through medio-lateral oblique (MLO) and bilateral craniocaudal (CC) for each breast to develop an attention module. The proposed AFIM model generates using spatial attention maps obtained from the identical image through other breasts to identify bilaterally uneven areas and&lt;/span&gt;&lt;span lang="EN-US"&gt;class activation map (CAM) generated from two similar breast images to emphasize the feature channels connected to a single lesion in a breast. AFIM model may easily be included in ResNet-style architectures to develop multi-view classification models.&lt;/span&gt;

Fast Nano-IR Hyperspectral Imaging Empowered by Large-Dataset-Free Miniaturized Spatial-Spectral Network.

The emerging field of nanoscale infrared (nano-IR) offers label-free molecular contrast, yet its imaging speed is limited by point-by-point traverse acquisition of a three-dimensional (3D) data cube. Here, we develop a spatial-spectral network (SS-Net), a miniaturized deep-learning model, together with compressive sampling to accelerate the nano-IR imaging. The compressive sampling is performed in both the spatial and spectral domains to accelerate the imaging process. The SS-Net is trained to learn the mapping from small nano-IR image patches to the corresponding spectra. With this elaborated mapping strategy, the training can be finished quickly within several minutes using the subsampled data, eliminating the need for a large-labeled dataset of common deep learning methods. We also designed an efficient loss function, which incorporates the image and spectral similarity to enhance the training. We first validate the SS-Net on an open stimulated Raman-scattering dataset; the results exhibit the potential of 10-fold imaging speed improvement with state-of-the-art performance. We then demonstrate the versatility of this approach on atomic force microscopy infrared (AFM-IR) microscopy with 7-fold imaging speed improvement, even on nanoscale Fourier transform infrared (nano-FTIR) microscopy with up to 261.6 folds faster imaging speed. We further showcase the generalization of this method on AFM-force volume-based multiparametric nanoimaging. This method establishes a paradigm for rapid nano-IR imaging, opening new possibilities for cutting-edge research in materials, photonics, and beyond.

Characterization of Rigid and Flexible Digital Detector Arrays Using ASTM E2597/E2597M-22

ASTM E2597/E2597M-22 is a standard that summarizes a standard practice for characterization of digital detector arrays (DDAs) by manufacturers. The standard consists of a series of tests intended to deliver quantitative results for a given model of detector’s characteristics as they should perform upon sale. It was written for use in characterizing rigid detectors, but the tests are applicable to bendable DDAs, utilized for single viewing. Bendable DDAs, unlike their rigid counterparts, are made to conform around parts being inspected, such as pipes, in a similar application to film and computed radiography for imaging curved objects. As bendable DDAs are made using the same scintillators as flat panel DDAs and have similar pixel pitch and resolution capabilities, it is expected that the results from most characterization tests should be similar. Conformable DDAs produce similar imaging results as rigid DDAs but have the flexibility of use for curved surfaces. They offer a valuable tool for customers currently using film and CR who are looking to expedite their inspections or move away from use of consumable products. We performed these characterization procedures on conformable DDA models and rigid DDA models utilizing proprietary INDUSTREX software developed by Carestream Health, Inc. The experimental results are presented as part of this paper.

Imaging the Male Breast: Gynecomastia, Male Breast Cancer, and Beyond.

The number of men undergoing breast imaging has increased in recent years, according to some reports. Most male breast concerns are related to benign causes, most commonly gynecomastia. The range of abnormalities typically encountered in the male breast is less broad than that encountered in women, given that lobule formation rarely occurs in men. Other benign causes of male breast palpable abnormalities with characteristic imaging findings include lipomas, sebaceous or epidermal inclusion cysts, and intramammary lymph nodes. Male breast cancer (MBC) is rare, representing up to 1% of breast cancer cases, but some data indicate that its incidence is increasing. MBC demonstrates some clinical features that overlap with those of gynecomastia, including a propensity for the subareolar breast. Men with breast cancer tend to present at a later stage than do women. MBC typically has similar imaging features to those of female breast cancer, often characterized by an irregular mass that may have associated calcifications. Occasionally, however, MBC has a benign-appearing imaging phenotype, with an oval shape and circumscribed margins, and therefore most solid breast masses in men require tissue diagnosis. Histopathologic evaluation may alternatively reveal other benign breast masses found in men, including papillomas, myofibroblastomas, and hemangiomas. Radiologists must be familiar with the breadth of male breast abnormalities to meet the rising challenge of caring for these patients. ©RSNA, 2024 Supplemental material is available for this article.

Diffeomorphic transformer-based abdomen MRI-CT deformable image registration.

Stereotactic body radiotherapy (SBRT) is a well-established treatment modality for liver metastases in patients unsuitable for surgery. Both CT and MRI are useful during treatment planning for accurate target delineation and to reduce potential organs-at-risk (OAR) toxicity from radiation. MRI-CT deformable image registration (DIR) is required to propagate the contours defined on high-contrast MRI to CT images. An accurate DIR method could lead to more precisely defined treatment volumes and superior OAR sparing on the treatment plan. Therefore, it is beneficial to develop an accurate MRI-CT DIR for liver SBRT. To create a new deep learning model that can estimate the deformation vector field (DVF) for directly registering abdominal MRI-CT images. The proposed method assumed a diffeomorphic deformation. By using topology-preserved deformation features extracted from the probabilistic diffeomorphic registration model, abdominal motion can be accurately obtained and utilized for DVF estimation. The model integrated Swin transformers, which have demonstrated superior performance in motion tracking, into the convolutional neural network (CNN) for deformation feature extraction. The model was optimized using a cross-modality image similarity loss and a surface matching loss. To compute the image loss, a modality-independent neighborhood descriptor (MIND) was used between the deformed MRI and CT images. The surface matching loss was determined by measuring the distance between the warped coordinates of the surfaces of contoured structures on the MRI and CT images. To evaluate the performance of the model, a retrospective study was carried out on a group of 50 liver cases that underwent rigid registration of MRI and CT scans. The deformed MRI image was assessed against the CT image using the target registration error (TRE), Dice similarity coefficient (DSC), and mean surface distance (MSD) between the deformed contours of the MRI image and manual contours of the CT image. When compared to only rigid registration, DIR with the proposed method resulted in an increase of the mean DSC values of the liver and portal vein from 0.850±0.102 and 0.628±0.129 to 0.903±0.044 and 0.763±0.073, a decrease of the mean MSD of the liver from 7.216±4.513mm to 3.232±1.483mm, and a decrease of the TRE from 26.238±2.769mm to 8.492±1.058mm. The proposed DIR method based on a diffeomorphic transformer provides an effective and efficient way to generate an accurate DVF from an MRI-CT image pair of the abdomen. It could be utilized in the current treatment planning workflow for liver SBRT.

Deep Learning Based Cystoscopy Image Enhancement.

Background: Endoscopy image enhancement technology provides doctors with clearer and more detailed images for observation and diagnosis, allowing doctors to assess lesions more accurately. Unlike most other endoscopy images, cystoscopy images face more complex and diverse image degradation because of their underwater imaging characteristics. Among the various causes of image degradation, the blood haze resulting from bladder mucosal bleeding make the background blurry and unclear, severely affecting diagnostic efficiency, even leading to misjudgment. Materials and Methods: We propose a deep learning-based approach to mitigate the impact of blood haze on cystoscopy images. The approach consists of two parts as follows: a blood haze removal network and a contrast enhancement algorithm. First, we adopt Feature Fusion Attention Network (FFA-Net) and transfer learning in the field of deep learning to remove blood haze from cystoscopy images and introduce perceptual loss to constrain the network for better visual results. Second, we enhance the image contrast by remapping the gray scale of the blood haze-free image and performing weighted fusion of the processed image and the original image. Results: In the blood haze removal stage, the algorithm proposed in this article achieves an average peak signal-to-noise ratio of 29.44 decibels, which is 15% higher than state-of-the-art traditional methods. The average structural similarity and perceptual image patch similarity reach 0.9269 and 0.1146, respectively, both superior to state-of-the-art traditional methods. Besides, our method is the best in keeping color balance after removing the blood haze. In the image enhancement stage, our algorithm enhances the contrast of vessels and tissues while preserving the original colors, expanding the dynamic range of the image. Conclusion: The deep learning-based cystoscopy image enhancement method is significantly better than other traditional methods in both qualitative and quantitative evaluation. The application of artificial intelligence will provide clearer, higher contrast cystoscopy images for medical diagnosis.

Enhancement of Mine Images through Reflectance Estimation of V Channel Using Retinex Theory

The dim lighting and excessive dust in underground mines often result in uneven illumination, blurriness, and loss of detail in surveillance images, which hinders subsequent intelligent image recognition. To address the limitations of the existing image enhancement algorithms in terms of generalization and accuracy, this paper proposes an unsupervised method for enhancing mine images in the hue–saturation–value (HSV) color space. Inspired by the HSV color space, the method first converts RGB images to the HSV space and integrates Retinex theory into the brightness (V channel). Additionally, a random perturbation technique is designed for the brightness. Within the same scene, a U-Net-based reflectance estimation network is constructed by enforcing consistency between the original reflectance and the perturbed reflectance, incorporating ResNeSt blocks and a multi-scale channel pixel attention module to improve accuracy. Finally, an enhanced image is obtained by recombining the original hue (H channel), brightness, and saturation (S channel), and converting back to the RGB space. Importantly, this image enhancement algorithm does not require any normally illuminated images during training. Extensive experiments demonstrated that the proposed method outperformed most existing unsupervised low-light image enhancement methods, qualitatively and quantitatively, achieving a competitive performance comparable to many supervised methods. Specifically, our method achieved the highest PSNR value of 22.18, indicating significant improvements compared to the other methods, and surpassing the second-best WCDM method by 10.3%. In terms of SSIM, our method also performed exceptionally well, achieving a value of 0.807, surpassing all other methods, and improving upon the second-place WCDM method by 19.5%. These results demonstrate that our proposed method significantly enhanced image quality and similarity, far exceeding the performance of the other algorithms.

Disambiguity and Alignment: An Effective Multi-Modal Alignment Method for Cross-Modal Recipe Retrieval.

As a prominent topic in food computing, cross-modal recipe retrieval has garnered substantial attention. However, the semantic alignment across food images and recipes cannot be further enhanced due to the lack of intra-modal alignment in existing solutions. Additionally, a critical issue named food image ambiguity is overlooked, which disrupts the convergence of models. To these ends, we propose a novel Multi-Modal Alignment Method for Cross-Modal Recipe Retrieval (MMACMR). To consider inter-modal and intra-modal alignment together, this method measures the ambiguous food image similarity under the guidance of their corresponding recipes. Additionally, we enhance recipe semantic representation learning by involving a cross-attention module between ingredients and instructions, which is effective in supporting food image similarity measurement. We conduct experiments on the challenging public dataset Recipe1M; as a result, our method outperforms several state-of-the-art methods in commonly used evaluation criteria.