Recently, iris recognition techniques have achieved great performance in identification. Among authentication techniques, iris recognition systems have received attention very much due to their rich iris texture which gives robust standards for identifying individuals. Notwithstanding this, there are several challenges in unrestricted recognition environments. In this article, the researchers present the techniques used in different phases of the recognition system of the iris image. The researchers also reviewed the methods associated with each phase. The recognition system is divided into seven phases, namely, the acquisition phase in which the iris images are acquired, the preprocessing phase in which the quality of the iris image is improved, the segmentation phase in which the iris region is separated from the background of the image, the normalization phase in which the segmented iris region is shaped into a rectangle, the feature extraction phase in which the features of the iris region are extracted, the feature selection phase in which the unique features of the iris are selected using feature selection techniques, and finally the classification phase in which the iris images are classified. This article also explains the two approaches of iris recognition which are the traditional approach and the deep learning approach. In addition, the researchers discuss the advantages and disadvantages of previous techniques as well as the limitations and benefits of both the traditional and deep learning approaches of iris recognition. This study can be considered as an initial step towards a large‐scale study about iris recognition.

Context: Routing algorithms play a crucial role in the performance of Network-on-Chip (NoC)-based Multi-Processor Systems-on-Chip (MPSoC). However, the selection of appropriate and effective routing algorithms poses a challenge for designers, given the multitude of assessment criteria, data fluctuations, and varying criteria importance. Objective: In this study, we propose a comprehensive assessment of various routing algorithms, aiming to identify the most suitable and effective routing algorithm that satisfies designers’ system-level requirements and assessment criteria. Method: This research integrates the Fuzzy-Weighted Zero-Inconsistency (FWZIC) method and the Fuzzy Decision by Opinion Score Method (FDOSM). The utilisation of the Z-Cloud Rough Numbers (ZCRNs) environment addresses the challenge of two types of uncertainty, providing a framework for managing ambiguity in the data and achieving a higher level of data freedom. Our methodology consists of two main phases. Firstly, the decision matrix is constructed based on the performance assessment criteria and routing algorithms. Secondly, we employ the ZCR-FWZIC method to derive the weights for each criterion and subsequently employ the ZCR-FDOSM-BM approach to rank the routing algorithms. Result: The analysis reveals that Adaptive Dimensional Bubble Routing (ADBR), Message-based Congestion-Aware Routing (MCAR), and Dynamic and Adaptive Routing Algorithm (DyAd) are ranked as the top three routing algorithms, respectively. Conclusion: This research presents essential implications for designers and system engineers involved in NoC-based MPSoC, offering insights to enhance decision-making processes and facilitate the selection of an appropriate routing algorithm.

Context: Network-on-Chips (NoCs) serve as essential interconnection infrastructures in Multi-processor System-on-Chip (MPSoC) designs, emphasizing flexibility, extensibility, and low power consumption. The effectiveness of communication within NoCs relies heavily on the routing algorithm employed. However, the routing process faces significant challenges, such as deadlock, livelock, congestion, and faults, which impact the Design Space Exploration (DSE) process. Objective: In this research, we propose a prediction model-based ANN with a Metaheuristic Optimization approach for predicting the utilized routing algorithm by the NoC-based MPSoC platform during the DSE in order to reduce the time required to specify the NoC-based MPSoC platform configurations. Method: The main idea of the proposed method is to develop a prediction model, specifically an Artificial Neural Network (ANN) optimized using the Guaranteed Convergence Arithmetic Optimization Algorithm (GCAOA-ANN), for predicting the utilized routing algorithm in NoC-based MPSoC platform during the DSE process. The methodology consists of two phases. Firstly, an automated strategy is proposed to execute and control the NOXIM simulator, enabling the simulation and collection of NoC-based MPSoC scenarios under different constraints, rules of NoC design (including routing protocols), and performance metrics. Consequently, NoC traffic data is gathered. Secondly, building upon the work of previous researchers, who have suggested the development of hybrid models to improve accuracy, this research introduces a novel methodology incorporating an ANN optimized using the GCAOA-ANN to predict the routing algorithm employed by the NoC-based MPSoC. Results: The hybrid GCAOA-ANN model demonstrated superior performance compared to other models, including PSOGW-ANN, SMA-ANN, AOA-ANN, Random Forest, and ANN. Where the learning rate and the number of nodes in the first and second hidden layers were: (0.0023, 1, 20) for the GCAOA-ANN model. In addition, the evaluation metrics of the GCAOA-ANN model showed excellent results, with an R2 value of (99.80%), an RMSE of (0.0000135 m3/s), an MSE of (0.0000118 m3/s), an SI of (0.000077), an MBE of (0.0000028), and a MAR of (0.000187616 m3/s). Conclusion: The outcomes of this research are anticipated to offer a guide for individuals interested in conducting research on On-Chip communication routing protocols and DSE in NoC-based MPSoCs in the context of IoT using ML approaches, while the GCAOA-ANN has the potential to generalize to new problems.

Deep Learning (DL) techniques have significantly improved the diagnostic accuracy in healthcare, particularly for detecting and classifying skin cancer. These advancements are expected to assist healthcare professionals in delivering more accurate, efficient, and timely diagnosis, ultimately improving patient outcomes and facilitating early detection and treatment of skin cancer. Additionally, medical imaging technologies, such as magnetic resonance imaging (MRI) and computed tomography (CT) scans remain critical tools for diagnosing dermatological conditions. However, interpreting these images can be challenging because of the overlapping structures and varying image quality. This study explored the application of DL in skin cancer diagnosis, with a particular focus on advances in image segmentation and classification. It reviews the development of DL-based models, including Convolutional Neural Networks (CNNs), and evaluates their effectiveness in skin lesion detection. This study examined the critical challenges associated with deploying DL models in clinical practice, including dataset diversity, model interpretability, and the feasibility of real-world implementation. It explores essential factors such as the selection of network architectures and data preprocessing techniques, emphasizing their influence on model performance. Additionally, this study identifies research gaps and suggests future directions for optimizing DL models for dermatological applications.

Iris segmentation is a significant phase in the iris recognition process because segmentation errors cascade into all subsequent phases. Therefore, it is important that errors in iris segmentation are minimised. The U-Net architecture that uses a deep learning approach was previously adopted for this task but its performance was affected by the deformation of iris images caused by various noise factors in unconstrained (non-ideal) environments. Scratches, blurriness, dirt, specular reflections and other noise factors are some of the challenges faced in unconstrained environments when eyeglasses are present in the original images. Additionally, the performance of the iris segmentation was degraded due to problems of exploding gradient or vanishing gradient and the loss of information. This paper proposes a multi-segmentation network called MS-Net, based on a deep learning approach, that aims to capture high-level semantic features while maintaining spatial information to improve the accuracy of iris segmentation. MS-Net consists of three principal segments: a feature encoder network, a multi-scale context feature extractor network (MSCFE-Net) and a feature decoder network. MSCFE-Net a multi-scale context feature extractor network is constructed from a dilated residual multi-convolutional network module and a pyramid pooling residual model based on an attention convolutional module. In addition, the proposed MS-Net contains dense connections within the feature decoder network to decrease training difficulty, by using only a few training samples. The accuracy of MS-Net was evaluated on the CASIA-Iris.V4-1000 and UBIRIS.V2 databases. The performance of our proposed MS-Net method on the CASIA-Iris.V4-1000 and UBIRIS.V2 databases achieved an overall accuracy of 97.11% and 96.128%, respectively. Experiment results show that MS-Net is able to achieve better results compared to earlier methods used for the same purpose.

In geology and mineralogy, optical microscopic images have become a primary research focus for intelligent mineral recognition due to their low equipment cost, ease of use, and distinct mineral characteristics in imaging. However, due to their close reflectivity or transparency, some minerals are not easily distinguished from other minerals or background. Secondly, the number of background pixels often vastly exceeds the number of pixels for individual mineral particles, and the number of pixels of different mineral particles in the image also varies significantly. These have led to the issue of data imbalance. This imbalance results in lower recognition accuracy for categories with fewer samples. To address these issues, a flexible ensemble learning for semantic segmentation based on multiple optimized Res-UNet models is proposed, introducing dice loss and focal loss functions and incorporating a pre-positioned spatial transformer networks block. Twelve optimized Res-UNet models were used to construct multiple Res-UNet ensemble learnings using heterogeneous ensemble strategies. The results demonstrate that the system integrated with five learners using the weighted voting fusion method (RUEL-5-WV) achieved the best performance with a mean Intersection over Union (mIOU) of 91.65 across all nine categories and an IOU of 84.33 for the transparent mineral (gangue). The results indicate that this ensemble learning scheme outperforms individual optimized Res-UNet models. Compared to the classical Deeplabv3 and PSPNet, this scheme also exhibits significant advantages.