Mulberry leaf dataset for image classification task.

This manuscript presents a mulberry leaf dataset collected from five provinces within three regions in Thailand. The dataset contains ten categories of mulberry leaves. We proposed this dataset due to the challenges of classifying leaf images taken in natural environments arising from high inter-class similarity and variations in illumination and background conditions (multiple leaves from a mulberry tree and shadows appearing in the leaf images). We highlight that our research team recorded mulberry leaves independently from various perspectives during our data acquisition using multiple camera types. The mulberry leaf dataset can serve as vital input data passed to computer vision algorithms (conventional deep learning and vision transformer algorithms) for creating image recognition systems. The dataset will allow other researchers to propose novel computer vision techniques to approach mulberry recognition challenges.

Evolutionary Strategies Enable Systematic and Reliable Uncertainty Quantification: A Proof-of-Concept Pilot Study on Resting-State Functional MRI Language Lateralization.

Reliable and trustworthy artificial intelligence (AI), particularly in high-stake medical diagnoses, necessitates effective uncertainty quantification (UQ). Existing UQ methods using model ensembles often introduce invalid variability or computational complexity, rendering them impractical and ineffective in clinical workflow. We propose a UQ approach based on deep neuroevolution (DNE), a data-efficient optimization strategy. Our goal is to replicate trends observed in expert-based UQ. We focused on language lateralization maps from resting-state functional MRI (rs-fMRI). Fifty rs-fMRI maps were divided into training/testing (30:20) sets, representing two labels: "left-dominant" and "co-dominant." DNE facilitated acquiring an ensemble of 100 models with high training and testing set accuracy. Model uncertainty was derived from distribution entropies over the 100 model predictions. Expert reviewers provided user-based uncertainties for comparison. Model (epistemic) and user-based (aleatoric) uncertainties were consistent in the independently and identically distributed (IID) testing set, mainly indicating low uncertainty. In a mostly out-of-distribution (OOD) holdout set, both model and user-based entropies correlated but displayed a bimodal distribution, with one peak representing low and another high uncertainty. We also found a statistically significant positive correlation between epistemic and aleatoric uncertainties. DNE-based UQ effectively mirrored user-based uncertainties, particularly highlighting increased uncertainty in OOD images. We conclude that DNE-based UQ correlates with expert assessments, making it reliable for our use case and potentially for other radiology applications.

SPMLD: A skin pathological image dataset for non-melanoma with detailed lesion area annotation.

Skin tumors are the most common tumors in humans and the clinical characteristics of three common non-melanoma tumors (IDN, SK, BCC) are similar, resulting in a high misdiagnosis rate. The accurate differential diagnosis of these tumors needs to be judged based on pathological images. However, a shortage of experienced dermatological pathologists leads to bias in the diagnostic accuracy of these skin tumors in China. In this paper, we establish a skin pathological image dataset, SPMLD, for three non-melanoma to achieve automatic and accurate intelligent identification for them. Meanwhile, we propose a lesion-area-based enhanced classification network with the KLS module and an attention module. Specifically, we first collect thousands of H&E-stained tissue sections from patients with clinically and pathologically confirmed IDN, SK, and BCC from a single-center hospital. Then, we scan them to construct a pathological image dataset of these three skin tumors. Furthermore, we mark the complete lesion area of the entire pathology image to better learn the pathologist's diagnosis process. In addition, we applied the proposed network for lesion classification prediction on the SPMLD dataset. Finally, we conduct a series of experiments to demonstrate that this annotation and our network can effectively improve the classification results of various networks. The source dataset and code are available at https://github.com/efss24/SPMLD.git.

An encoding framework for binarized images using hyperdimensional computing.

Introduction: Hyperdimensional Computing (HDC) is a brain-inspired and lightweight machine learning method. It has received significant attention in the literature as a candidate to be applied in the wearable Internet of Things, near-sensor artificial intelligence applications, and on-device processing. HDC is computationally less complex than traditional deep learning algorithms and typically achieves moderate to good classification performance. A key aspect that determines the performance of HDC is encoding the input data to the hyperdimensional (HD) space. Methods: This article proposes a novel lightweight approach relying only on native HD arithmetic vector operations to encode binarized images that preserves the similarity of patterns at nearby locations by using point of interest selection and local linear mapping. Results: The method reaches an accuracy of 97.92% on the test set for the MNIST data set and 84.62% for the Fashion-MNIST data set. Discussion: These results outperform other studies using native HDC with different encoding approaches and are on par with more complex hybrid HDC models and lightweight binarized neural networks. The proposed encoding approach also demonstrates higher robustness to noise and blur compared to the baseline encoding.

DP-SSLoRA: A privacy-preserving medical classification model combining differential privacy with self-supervised low-rank adaptation.

BACKGROUND AND OBJECTIVE: Concerns about patient privacy issues have limited the application of medical deep learning models in certain real-world scenarios. Differential privacy (DP) can alleviate this problem by injecting random noise into the model. However, naively applying DP to medical models will not achieve a satisfactory balance between privacy and utility due to the high dimensionality of medical models and the limited labeled samples. METHODS: This work proposed the DP-SSLoRA model, a privacy-preserving classification model for medical images combining differential privacy with self-supervised low-rank adaptation. In this work, a self-supervised pre-training method is used to obtain enhanced representations from unlabeled publicly available medical data. Then, a low-rank decomposition method is employed to mitigate the impact of differentially private noise and combined with pre-trained features to conduct the classification task on private datasets. RESULTS: In the classification experiments using three real chest-X ray datasets, DP-SSLoRA achieves good performance with strong privacy guarantees. Under the premise of ɛ=2, with the AUC of 0.942 in RSNA, the AUC of 0.9658 in Covid-QU-mini, and the AUC of 0.9886 in Chest X-ray 15k. CONCLUSION: Extensive experiments on real chest X-ray datasets show that DP-SSLoRA can achieve satisfactory performance with stronger privacy guarantees. This study provides guidance for studying privacy-preserving in the medical field. Source code is publicly available online. https://github.com/oneheartforone/DP-SSLoRA.

MetaV: A Pioneer in feature Augmented Meta-Learning Based Vision Transformer for Medical Image Classification.

Image classification, a fundamental task in computer vision, faces challenges concerning limited data handling, interpretability, improved feature representation, efficiency across diverse image types, and processing noisy data. Conventional architectural approaches have made insufficient progress in addressing these challenges, necessitating architectures capable of fine-grained classification, enhanced accuracy, and superior generalization. Among these, the vision transformer emerges as a noteworthy computer vision architecture. However, its reliance on substantial data for training poses a drawback due to its complexity and high data requirements. To surmount these challenges, this paper proposes an innovative approach, MetaV, integrating meta-learning into a vision transformer for medical image classification. N-way K-shot learning is employed to train the model, drawing inspiration from human learning mechanisms utilizing past knowledge. Additionally, deformational convolution and patch merging techniques are incorporated into the vision transformer model to mitigate complexity and overfitting while enhancing feature representation. Augmentation methods such as perturbation and Grid Mask are introduced to address the scarcity and noise in medical images, particularly for rare diseases. The proposed model is evaluated using diverse datasets including Break His, ISIC 2019, SIPaKMed, and STARE. The achieved performance accuracies of 89.89%, 87.33%, 94.55%, and 80.22% for Break His, ISIC 2019, SIPaKMed, and STARE, respectively, present evidence validating the superior performance of the proposed model in comparison to conventional models, setting a new benchmark for meta-vision image classification models.

When an extra rejection class meets out-of-distribution detection in long-tailed image classification.

Detecting Out-of-Distribution (OOD) inputs is essential for reliable deep learning in the open world. However, most existing OOD detection methods have been developed based on training sets that exhibit balanced class distributions, making them susceptible when confronted with training sets following a long-tailed distribution. To alleviate this problem, we propose an effective three-branch training framework, which demonstrates the efficacy of incorporating an extra rejection class along with auxiliary outlier training data for effective OOD detection in long-tailed image classification. In our proposed framework, all outlier training samples are assigned the label of the rejection class. We employ an inlier loss, an outlier loss, and a Tail-class prototype induced Supervised Contrastive Loss (TSCL) to train both the in-distribution classifier and OOD detector within one network. During inference, the OOD detector is constructed using the rejection class. Extensive experimental results demonstrate that the superior OOD detection performance of our proposed method in long-tailed image classification. For example, in the more challenging case where CIFAR100-LT is used as in-distribution, our method improves the average AUROC by 1.23% and reduces the average FPR95 by 3.18% compared to the baseline method utilizing Outlier Exposure (OE). Code is available at github.

Modified osprey algorithm for optimizing capsule neural network in leukemia image recognition.

The diagnosis of leukemia is a serious matter that requires immediate and accurate attention. This research presents a revolutionary method for diagnosing leukemia using a Capsule Neural Network (CapsNet) with an optimized design. CapsNet is a cutting-edge neural network that effectively captures complex features and spatial relationships within images. To improve the CapsNet's performance, a Modified Version of Osprey Optimization Algorithm (MOA) has been utilized. Thesuggested approach has been tested on the ALL-IDB database, a widely recognized dataset for leukemia image classification. Comparative analysis with various machine learning techniques, including Combined combine MobilenetV2 and ResNet18 (MBV2/Res) network, Depth-wise convolution model, a hybrid model that combines a genetic algorithm with ResNet-50V2 (ResNet/GA), and SVM/JAYA demonstrated the superiority of our method in different terms. As a result, the proposed method is a robust and powerful tool for diagnosing leukemia from medical images.

Characterization of Heart Diseases per Single Lead Using ECG Images and CNN-2D.

Cardiopathy has become one of the predominant global causes of death. The timely identification of different types of heart diseases significantly diminishes mortality risk and enhances the efficacy of treatment. However, fast and efficient recognition necessitates continuous monitoring, encompassing not only specific clinical conditions but also diverse lifestyles. Consequently, an increasing number of studies are striving to automate and progress in the identification of different cardiopathies. Notably, the assessment of electrocardiograms (ECGs) is crucial, given that it serves as the initial diagnostic test for patients, proving to be both the simplest and the most cost-effective tool. This research employs a customized architecture of Convolutional Neural Network (CNN) to forecast heart diseases by analyzing the images of both three bands of electrodes and of each single electrode signal of the ECG derived from four distinct patient categories, representing three heart-related conditions as well as a spectrum of healthy controls. The analyses are conducted on a real dataset, providing noteworthy performance (recall greater than 80% for the majority of the considered diseases and sometimes even equal to 100%) as well as a certain degree of interpretability thanks to the understanding of the importance a band of electrodes or even a single ECG electrode can have in detecting a specific heart-related pathology.

NTCE-KD: Non-Target-Class-Enhanced Knowledge Distillation.

Most logit-based knowledge distillation methods transfer soft labels from the teacher model to the student model via Kullback-Leibler divergence based on softmax, an exponential normalization function. However, this exponential nature of softmax tends to prioritize the largest class (target class) while neglecting smaller ones (non-target classes), leading to an oversight of the non-target classes's significance. To address this issue, we propose Non-Target-Class-Enhanced Knowledge Distillation (NTCE-KD) to amplify the role of non-target classes both in terms of magnitude and diversity. Specifically, we present a magnitude-enhanced Kullback-Leibler (MKL) divergence multi-shrinking the target class to enhance the impact of non-target classes in terms of magnitude. Additionally, to enrich the diversity of non-target classes, we introduce a diversity-based data augmentation strategy (DDA), further enhancing overall performance. Extensive experimental results on the CIFAR-100 and ImageNet-1k datasets demonstrate that non-target classes are of great significance and that our method achieves state-of-the-art performance across a wide range of teacher-student pairs.

Proto-Adapter: Efficient Training-Free CLIP-Adapter for Few-Shot Image Classification.

Large vision-language models, such as Contrastive Vision-Language Pre-training (CLIP), pre-trained on large-scale image-text datasets, have demonstrated robust zero-shot transfer capabilities across various downstream tasks. To further enhance the few-shot recognition performance of CLIP, Tip-Adapter augments the CLIP model with an adapter that incorporates a key-value cache model constructed from the few-shot training set. This approach enables training-free adaptation and has shown significant improvements in few-shot recognition, especially with additional fine-tuning. However, the size of the adapter increases in proportion to the number of training samples, making it difficult to deploy in practical applications. In this paper, we propose a novel CLIP adaptation method, named Proto-Adapter, which employs a single-layer adapter of constant size regardless of the amount of training data and even outperforms Tip-Adapter. Proto-Adapter constructs the adapter's weights based on prototype representations for each class. By aggregating the features of the training samples, it successfully reduces the size of the adapter without compromising performance. Moreover, the performance of the model can be further enhanced by fine-tuning the adapter's weights using a distance margin penalty, which imposes additional inter-class discrepancy to the output logits. We posit that this training scheme allows us to obtain a model with a discriminative decision boundary even when trained with a limited amount of data. We demonstrate the effectiveness of the proposed method through extensive experiments of few-shot classification on diverse datasets.

Energy-Efficient Edge and Cloud Image Classification with Multi-Reservoir Echo State Network and Data Processing Units.

In an era dominated by Internet of Things (IoT) devices, software-as-a-service (SaaS) platforms, and rapid advances in cloud and edge computing, the demand for efficient and lightweight models suitable for resource-constrained devices such as data processing units (DPUs) has surged. Traditional deep learning models, such as convolutional neural networks (CNNs), pose significant computational and memory challenges, limiting their use in resource-constrained environments. Echo State Networks (ESNs), based on reservoir computing principles, offer a promising alternative with reduced computational complexity and shorter training times. This study explores the applicability of ESN-based architectures in image classification and weather forecasting tasks, using benchmarks such as the MNIST, FashionMnist, and CloudCast datasets. Through comprehensive evaluations, the Multi-Reservoir ESN (MRESN) architecture emerges as a standout performer, demonstrating its potential for deployment on DPUs or home stations. In exploiting the dynamic adaptability of MRESN to changing input signals, such as weather forecasts, continuous on-device training becomes feasible, eliminating the need for static pre-trained models. Our results highlight the importance of lightweight models such as MRESN in cloud and edge computing applications where efficiency and sustainability are paramount. This study contributes to the advancement of efficient computing practices by providing novel insights into the performance and versatility of MRESN architectures. By facilitating the adoption of lightweight models in resource-constrained environments, our research provides a viable alternative for improved efficiency and scalability in modern computing paradigms.

A universal multiple instance learning framework for whole slide image analysis.

BACKGROUND: The emergence of digital whole slide image (WSI) has driven the development of computational pathology. However, obtaining patch-level annotations is challenging and time-consuming due to the high resolution of WSI, which limits the applicability of fully supervised methods. We aim to address the challenges related to patch-level annotations. METHODS: We propose a universal framework for weakly supervised WSI analysis based on Multiple Instance Learning (MIL). To achieve effective aggregation of instance features, we design a feature aggregation module from multiple dimensions by considering feature distribution, instances correlation and instance-level evaluation. First, we implement instance-level standardization layer and deep projection unit to improve the separation of instances in the feature space. Then, a self-attention mechanism is employed to explore dependencies between instances. Additionally, an instance-level pseudo-label evaluation method is introduced to enhance the available information during the weak supervision process. Finally, a bag-level classifier is used to obtain preliminary WSI classification results. To achieve even more accurate WSI label predictions, we have designed a key instance selection module that strengthens the learning of local features for instances. Combining the results from both modules leads to an improvement in WSI prediction accuracy. RESULTS: Experiments conducted on Camelyon16, TCGA-NSCLC, SICAPv2, PANDA and classical MIL benchmark datasets demonstrate that our proposed method achieves a competitive performance compared to some recent methods, with maximum improvement of 14.6 % in terms of classification accuracy. CONCLUSION: Our method can improve the classification accuracy of whole slide images in a weakly supervised way, and more accurately detect lesion areas.

UKSSL: Underlying Knowledge Based Semi-Supervised Learning for Medical Image Classification.

Goal: Deep learning techniques have made significant progress in medical image analysis. However, obtaining ground truth labels for unlabeled medical images is challenging as they often outnumber labeled images. Thus, training a high-performance model with limited labeled data has become a crucial challenge. Methods: This study introduces an underlying knowledge-based semi-supervised framework called UKSSL, consisting of two components: MedCLR extracts feature representations from the unlabeled dataset; UKMLP utilizes the representation and fine-tunes it with the limited labeled dataset to classify the medical images. Results: UKSSL evaluates on the LC25000 and BCCD datasets, using only 50% labeled data. It gets precision, recall, F1-score, and accuracy of 98.9% on LC25000 and 94.3%, 94.5%, 94.3%, and 94.1% on BCCD, respectively. These results outperform other supervised-learning methods using 100% labeled data. Conclusions: The UKSSL can efficiently extract underlying knowledge from the unlabeled dataset and perform better using limited labeled medical images.

KLSANet: Key local semantic alignment Network for few-shot image classification.

Few-shot image classification involves recognizing new classes with a limited number of labeled samples. Current local descriptor-based methods, while leveraging consistent low-level features across visible and invisible classes, face challenges including redundant adjacent information, irrelevant partial representation, and limited interpretability. This paper proposes KLSANet, a few-shot image classification approach based on key local semantic alignment network, which aligns key local semantics for accurate classification. Furthermore, we introduce a key local screening module to mitigate the influence of semantically irrelevant image parts on classification. KLSANet demonstrates superior performance on three benchmark datasets (CUB, Stanford Dogs, Stanford Cars), outperforming state-of-the-art methods in 1-shot and 5-shot settings with average improvements of 3.95% and 2.56% respectively. Visualization experiments demonstrate the interpretability of KLSANet predictions. Code is available at: https://github.com/ZitZhengWang/KLSANet.

Stable tensor neural networks for efficient deep learning.

Learning from complex, multidimensional data has become central to computational mathematics, and among the most successful high-dimensional function approximators are deep neural networks (DNNs). Training DNNs is posed as an optimization problem to learn network weights or parameters that well-approximate a mapping from input to target data. Multiway data or tensors arise naturally in myriad ways in deep learning, in particular as input data and as high-dimensional weights and features extracted by the network, with the latter often being a bottleneck in terms of speed and memory. In this work, we leverage tensor representations and processing to efficiently parameterize DNNs when learning from high-dimensional data. We propose tensor neural networks (t-NNs), a natural extension of traditional fully-connected networks, that can be trained efficiently in a reduced, yet more powerful parameter space. Our t-NNs are built upon matrix-mimetic tensor-tensor products, which retain algebraic properties of matrix multiplication while capturing high-dimensional correlations. Mimeticity enables t-NNs to inherit desirable properties of modern DNN architectures. We exemplify this by extending recent work on stable neural networks, which interpret DNNs as discretizations of differential equations, to our multidimensional framework. We provide empirical evidence of the parametric advantages of t-NNs on dimensionality reduction using autoencoders and classification using fully-connected and stable variants on benchmark imaging datasets MNIST and CIFAR-10.

Binary classification of dead detector elements in flat panel detectors using convolutional neural networks.

Objective.Medical physicists routinely perform quality assurance on digital detection systems, part of which involves the testing of flat panel detectors. Flat panels may degrade over time as an increasing number of individual detector elements begin to malfunction. The pixels that correspond to these elements are corrected for using information elsewhere in the detector system, however these corrected elements still constitute a loss in image quality for the system as a whole. These correction methods, as well as the location and number of dead detector elements, are often only available to the vendor of the digital detection system, but not to the medical physicist responsible for the quality assurance of the system.Approach.We greatly expand upon a previous work by providing a novel technique for classifying dead detector elements at single pixel resolution. We also demonstrate that this technique can be trained on one detector, and then tested and validated on another with moderate success, which demonstrates some ability to generalize to different detectors. The technique requires 3 flat field, or 'noise', images to be taken to predict the dead detector element maps for the system.Main results.Models using only for-processing pixel data were unable to successfully generalize from one detector to the other. Models preprocessed using the standard deviation across three for-processing images were able to classify dead detector element maps with an F1score ranging from 0.4527 to 0.8107 and recall ranging from 0.5420 to 0.9303 with better performance, on average, observed using the low exposure data set.Significance. Many physicists do not have access to the dead detector maps for their diagnostic digital radiography systems. CNNs are capable of predicting the dead detector maps of flat panel detectors with single pixel resolution. Physicists can implement this tool by acquiring three flat field images and then inputting them into the model. Model performance saw a marginal increase when trained on the low exposure set data, as opposed to the high exposure set data, indicating high exposure, low relative noise images may not be necessary for optimal performance. Model performance across detectors manufactured by different vendors requires further investigation.

Histological tissue classification with a novel statistical filter-based convolutional neural network.

Deep networks have been of considerable interest in literature and have enabled the solution of recent real-world applications. Due to filters that offer feature extraction, Convolutional Neural Network (CNN) is recognized as an accurate, efficient and trustworthy deep learning technique for the solution of image-based challenges. The high-performing CNNs are computationally demanding even if they produce good results in a variety of applications. This is because a large number of parameters limit their ability to be reused on central processing units with low performance. To address these limitations, we suggest a novel statistical filter-based CNN (HistStatCNN) for image classification. The convolution kernels of the designed CNN model were initialized by continuous statistical methods. The performance of the proposed filter initialization approach was evaluated on a novel histological dataset and various histopathological benchmark datasets. To prove the efficiency of statistical filters, three unique parameter sets and a mixed parameter set of statistical filters were applied to the designed CNN model for the classification task. According to the results, the accuracy of GoogleNet, ResNet18, ResNet50 and ResNet101 models were 85.56%, 85.24%, 83.59% and 83.79%, respectively. The accuracy was improved by 87.13% by HistStatCNN for the histological data classification task. Moreover, the performance of the proposed filter generation approach was proved by testing on various histopathological benchmark datasets, increasing average accuracy rates. Experimental results validate that the proposed statistical filters enhance the performance of the network with more simple CNN models.

Convolutional neural network application for supply-demand matching in Zhuang ethnic clothing image classification.

This study aims to design a classification technique suitable for Zhuang ethnic clothing images by integrating the concept of supply-demand matching and convolutional neural networks. Firstly, addressing the complex structure and unique visual style of Zhuang ethnic clothing, this study proposes an image resolution model based on supply-demand matching and convolutional networks. By integrating visual style and label constraints, this model accurately extracts local features. Secondly, the model's effectiveness and resolution performance are analyzed through various performance metrics in experiments. The results indicate a significant improvement in detection accuracy at different annotation points. The model outperforms other comparative methods in pixel accuracy (90.5%), average precision (83.7%), average recall (80.1%), and average F1 score (81.2%). Next, this study introduces a clothing image classification algorithm based on key points and channel attention. Through key point detection and channel attention mechanisms, image features are optimized, enabling accurate classification and attribute prediction of Zhuang ethnic clothing. Experimental results demonstrate a notable enhancement in category classification and attribute prediction, with classification accuracy and recall exceeding 90% in top-k tasks, showcasing outstanding performance. In conclusion, this study provides innovative approaches and effective solutions for deep learning classification of Zhuang ethnic clothing images.

Mapping of uncontrolled dumpsites in arid regions through remote sensing and image processing.

Due to increased urbanization, the development of new areas, construction of new houses and buildings and uncontrolled dumpsites (UDSs) are becoming a challenge facing local authorities in Saudi Arabia. UDSs pose health risks to the public, potentially deteriorating the environment around them and reducing the value of ongoing development areas. The local municipalities rely on field surveys and citizen reports. This can be inefficient because UDSs are often discovered too late, and remediating them can be costly. This study aimed to assess the conditions of UDSs in two cities in the Eastern Province of Saudi Arabia, Dammam and Hafer Al-Batin, using satellite image classification assessment techniques. The assessment included mapping the UDS locations and studying the spectral reflectance of the materials found in these dumpsites. The study provided a mapping of 62 UDS locations totalling around 13.01 km2 in the broader study area. UDS detections using remote sensing were followed by ground truthing and in situ measurements using a spectroradiometer. In addition, the spectral reflectance of 21 commonly deposited UDS materials was studied, and a spectral library was created for these materials for future use by local authorities.