Explainable Artificial Intelligence(XAI)Model for Cancer Image Classification

The use of Explainable Artificial Intelligence(XAI)models becomes increasingly important for making decisions in smart healthcare environments.It is to make sure that decisions are based on trustworthy algorithms and that healthcare workers understand the decisions made by these algorithms.These models can potentially enhance interpretability and explainability in decision-making processes that rely on artificial intelligence.Nevertheless,the intricate nature of the healthcare field necessitates the utilization of sophisticated models to classify cancer images.This research presents an advanced investigation of XAI models to classify cancer images.It describes the different levels of explainability and interpretability associated with XAI models and the challenges faced in deploying them in healthcare applications.In addition,this study proposes a novel framework for cancer image classification that incorporates XAI models with deep learning and advanced medical imaging techniques.The proposed model integrates several techniques,including end-to-end explainable evaluation,rule-based explanation,and useradaptive explanation.The proposed XAI reaches 97.72%accuracy,90.72%precision,93.72%recall,96.72%F1-score,9.55%FDR,9.66%FOR,and 91.18%DOR.It will discuss the potential applications of the proposed XAI models in the smart healthcare environment.It will help ensure trust and accountability in AI-based decisions,which is essential for achieving a safe and reliable smart healthcare environment.

Evaluating Privacy Leakage and Memorization Attacks on Large Language Models (LLMs) in Generative AI Applications

The recent interest in the deployment of Generative AI applications that use large language models (LLMs) has brought to the forefront significant privacy concerns, notably the leakage of Personally Identifiable Information (PII) and other confidential or protected information that may have been memorized during training, specifically during a fine-tuning or customization process. We describe different black-box attacks from potential adversaries and study their impact on the amount and type of information that may be recovered from commonly used and deployed LLMs. Our research investigates the relationship between PII leakage, memorization, and factors such as model size, architecture, and the nature of attacks employed. The study utilizes two broad categories of attacks: PII leakage-focused attacks (auto-completion and extraction attacks) and memorization-focused attacks (various membership inference attacks). The findings from these investigations are quantified using an array of evaluative metrics, providing a detailed understanding of LLM vulnerabilities and the effectiveness of different attacks.

Robust Facial Biometric Authentication System Using Pupillary Light Reflex for Liveness Detection of Facial Images

Pupil dynamics are the important characteristics of face spoofing detection.The face recognition system is one of the most used biometrics for authenticating individual identity.The main threats to the facial recognition system are different types of presentation attacks like print attacks,3D mask attacks,replay attacks,etc.The proposed model uses pupil characteristics for liveness detection during the authentication process.The pupillary light reflex is an involuntary reaction controlling the pupil’s diameter at different light intensities.The proposed framework consists of two-phase methodologies.In the first phase,the pupil’s diameter is calculated by applying stimulus(light)in one eye of the subject and calculating the constriction of the pupil size on both eyes in different video frames.The above measurement is converted into feature space using Kohn and Clynes model-defined parameters.The Support Vector Machine is used to classify legitimate subjects when the diameter change is normal(or when the eye is alive)or illegitimate subjects when there is no change or abnormal oscillations of pupil behavior due to the presence of printed photograph,video,or 3D mask of the subject in front of the camera.In the second phase,we perform the facial recognition process.Scale-invariant feature transform(SIFT)is used to find the features from the facial images,with each feature having a size of a 128-dimensional vector.These features are scale,rotation,and orientation invariant and are used for recognizing facial images.The brute force matching algorithm is used for matching features of two different images.The threshold value we considered is 0.08 for good matches.To analyze the performance of the framework,we tested our model in two Face antispoofing datasets named Replay attack datasets and CASIA-SURF datasets,which were used because they contain the videos of the subjects in each sample having three modalities(RGB,IR,Depth).The CASIA-SURF datasets showed an 89.9%Equal Error Rate,while the Replay Attack datasets showed a 92.1%Equal Error Rate.

Protecting LLMs against Privacy Attacks While Preserving Utility

The recent interest in the deployment of Generative AI applications that use large language models (LLMs) has brought to the forefront significant privacy concerns, notably the leakage of Personally Identifiable Information (PII) and other confidential or protected information that may have been memorized during training, specifically during a fine-tuning or customization process. This inadvertent leakage of sensitive information typically occurs when the models are subjected to black-box attacks. To address the growing concerns of safeguarding private and sensitive information while simultaneously preserving its utility, we analyze the performance of Targeted Catastrophic Forgetting (TCF). TCF involves preserving targeted pieces of sensitive information within datasets through an iterative pipeline which significantly reduces the likelihood of such information being leaked or reproduced by the model during black-box attacks, such as the autocompletion attack in our case. The experiments conducted using TCF evidently demonstrate its capability to reduce the extraction of PII while still preserving the context and utility of the target application.

GUARDIAN: A Multi-Tiered Defense Architecture for Thwarting Prompt Injection Attacks on LLMs

This paper introduces a novel multi-tiered defense architecture to protect language models from adversarial prompt attacks. We construct adversarial prompts using strategies like role emulation and manipulative assistance to simulate real threats. We introduce a comprehensive, multi-tiered defense framework named GUARDIAN (Guardrails for Upholding Ethics in Language Models) comprising a system prompt filter, pre-processing filter leveraging a toxic classifier and ethical prompt generator, and pre-display filter using the model itself for output screening. Extensive testing on Meta’s Llama-2 model demonstrates the capability to block 100% of attack prompts. The approach also auto-suggests safer prompt alternatives, thereby bolstering language model security. Quantitatively evaluated defense layers and an ethical substitution mechanism represent key innovations to counter sophisticated attacks. The integrated methodology not only fortifies smaller LLMs against emerging cyber threats but also guides the broader application of LLMs in a secure and ethical manner.

Smaller & Smarter: Score-Driven Network Chaining of Smaller Language Models

With the continuous evolution and expanding applications of Large Language Models (LLMs), there has been a noticeable surge in the size of the emerging models. It is not solely the growth in model size, primarily measured by the number of parameters, but also the subsequent escalation in computational demands, hardware and software prerequisites for training, all culminating in a substantial financial investment as well. In this paper, we present novel techniques like supervision, parallelization, and scoring functions to get better results out of chains of smaller language models, rather than relying solely on scaling up model size. Firstly, we propose an approach to quantify the performance of a Smaller Language Models (SLM) by introducing a corresponding supervisor model that incrementally corrects the encountered errors. Secondly, we propose an approach to utilize two smaller language models (in a network) performing the same task and retrieving the best relevant output from the two, ensuring peak performance for a specific task. Experimental evaluations establish the quantitative accuracy improvements on financial reasoning and arithmetic calculation tasks from utilizing techniques like supervisor models (in a network of model scenario), threshold scoring and parallel processing over a baseline study.

Whispered Tuning: Data Privacy Preservation in Fine-Tuning LLMs through Differential Privacy

The proliferation of Large Language Models (LLMs) across various sectors underscored the urgency of addressing potential privacy breaches. Vulnerabilities, such as prompt injection attacks and other adversarial tactics, could make these models inadvertently disclose their training data. Such disclosures could compromise personal identifiable information, posing significant privacy risks. In this paper, we proposed a novel multi-faceted approach called Whispered Tuning to address privacy leaks in large language models (LLMs). We integrated a PII redaction model, differential privacy techniques, and an output filter into the LLM fine-tuning process to enhance confidentiality. Additionally, we introduced novel ideas like the Epsilon Dial for adjustable privacy budgeting for differentiated Training Phases per data handler role. Through empirical validation, including attacks on non-private models, we demonstrated the robustness of our proposed solution SecureNLP in safeguarding privacy without compromising utility. This pioneering methodology significantly fortified LLMs against privacy infringements, enabling responsible adoption across sectors.

Learning Noise-Assisted Robust Image Features for Fine-Grained Image Retrieval

Fine-grained image search is one of the most challenging tasks in computer vision that aims to retrieve similar images at the fine-grained level for a given query image.The key objective is to learn discriminative fine-grained features by training deep models such that similar images are clustered,and dissimilar images are separated in the low embedding space.Previous works primarily focused on defining local structure loss functions like triplet loss,pairwise loss,etc.However,training via these approaches takes a long training time,and they have poor accuracy.Additionally,representations learned through it tend to tighten up in the embedded space and lose generalizability to unseen classes.This paper proposes a noise-assisted representation learning method for fine-grained image retrieval to mitigate these issues.In the proposed work,class manifold learning is performed in which positive pairs are created with noise insertion operation instead of tightening class clusters.And other instances are treated as negatives within the same cluster.Then a loss function is defined to penalize when the distance between instances of the same class becomes too small relative to the noise pair in that class in embedded space.The proposed approach is validated on CARS-196 and CUB-200 datasets and achieved better retrieval results(85.38%recall@1 for CARS-196%and 70.13%recall@1 for CUB-200)compared to other existing methods.

Generative Adversarial Network Based Approach towards Synthetically Generating Insider Threat Scenarios

This research paper explores the use of Generative Adversarial Networks (GANs) to synthetically generate insider threat scenarios. Insider threats pose significant risks to IT infrastructures, requiring effective detection and mitigation strategies. By training GAN models on historical insider threat data, synthetic scenarios resembling real-world incidents can be generated, including various tactics and procedures employed by insiders. The paper discusses the benefits, challenges, and ethical considerations associated with using GAN-generated data. The findings highlight the potential of GANs in enhancing insider threat detection and response capabilities, empowering organizations to fortify their defenses and proactively mitigate risks posed by internal actors.

Object Detection Meets LLMs: Model Fusion for Safety and Security

This paper proposes a novel model fusion approach to enhance predictive capabilities of vision and language models by strategically integrating object detection and large language models. We have named this multimodal integration approach as VOLTRON (Vision Object Linguistic Translation for Responsive Observation and Narration). VOLTRON is aimed at improving responses for self-driving vehicles in detecting small objects crossing roads and identifying merged or narrower lanes. The models are fused using a single layer to provide LLaMA2 (Large Language Model Meta AI) with object detection probabilities from YoloV8-n (You Only Look Once) translated into sentences. Experiments using specialized datasets showed accuracy improvements up to 88.16%. We provide a comprehensive exploration of the theoretical aspects that inform our model fusion approach, detailing the fundamental principles upon which it is built. Moreover, we elucidate the intricacies of the methodologies employed for merging these two disparate models, shedding light on the techniques and strategies used.

Screening of COVID-19 Patients Using Deep Learning and IoT Framework

In March 2020,the World Health Organization declared the coronavirus disease(COVID-19)outbreak as a pandemic due to its uncontrolled global spread.Reverse transcription polymerase chain reaction is a laboratory test that is widely used for the diagnosis of this deadly disease.However,the limited availability of testing kits and qualified staff and the drastically increasing number of cases have hampered massive testing.To handle COVID19 testing problems,we apply the Internet of Things and artificial intelligence to achieve self-adaptive,secure,and fast resource allocation,real-time tracking,remote screening,and patient monitoring.In addition,we implement a cloud platform for efficient spectrum utilization.Thus,we propose a cloudbased intelligent system for remote COVID-19 screening using cognitiveradio-based Internet of Things and deep learning.Specifically,a deep learning technique recognizes radiographic patterns in chest computed tomography(CT)scans.To this end,contrast-limited adaptive histogram equalization is applied to an input CT scan followed by bilateral filtering to enhance the spatial quality.The image quality assessment of the CT scan is performed using the blind/referenceless image spatial quality evaluator.Then,a deep transfer learning model,VGG-16,is trained to diagnose a suspected CT scan as either COVID-19 positive or negative.Experimental results demonstrate that the proposed VGG-16 model outperforms existing COVID-19 screening models regarding accuracy,sensitivity,and specificity.The results obtained from the proposed system can be verified by doctors and sent to remote places through the Internet.

Semantic Information Extraction from Multi-Corpora Using Deep Learning

Information extraction plays a vital role in natural language processing,to extract named entities and events from unstructured data.Due to the exponential data growth in the agricultural sector,extracting significant information has become a challenging task.Though existing deep learningbased techniques have been applied in smart agriculture for crop cultivation,crop disease detection,weed removal,and yield production,still it is difficult to find the semantics between extracted information due to unswerving effects of weather,soil,pest,and fertilizer data.This paper consists of two parts.An initial phase,which proposes a data preprocessing technique for removal of ambiguity in input corpora,and the second phase proposes a novel deep learning-based long short-term memory with rectification in Adam optimizer andmultilayer perceptron to find agricultural-based named entity recognition,events,and relations between them.The proposed algorithm has been trained and tested on four input corpora i.e.,agriculture,weather,soil,and pest&fertilizers.The experimental results have been compared with existing techniques and itwas observed that the proposed algorithm outperformsWeighted-SOM,LSTM+RAO,PLR-DBN,KNN,and Na飗e Bayes on standard parameters like accuracy,sensitivity,and specificity.

Novel integration of extreme learning machine and improved Harris hawks optimization with particle swarm optimization-based mutation for predicting soil consolidation parameter

The study proposes an improved Harris hawks optimization(IHHO) algorithm by integrating the standard Harris hawks optimization(HHO) algorithm and mutation-based search mechanism for developing a high-performance machine learning solution for predicting soil compression index. HHO is a newly introduced meta-heuristic optimization algorithm(MOA) used to solve continuous search problems.Compared to the original HHO, the proposed IHHO can evade trapping in local optima, which in turn raises the search capabilities and enhances the search mechanism relying on mutation. Subsequently, a novel meta-heuristic-based soft computing technique called ELM-IHHO was established by integrating IHHO and extreme learning machine(ELM) to estimate soil compression index. A sum of 688 consolidation test data was collected for this purpose from an ongoing dedicated freight corridor railway project. To evaluate the generalization capability of the proposed ELM-IHHO model, a detailed comparison between ELM-IHHO and other well-established MOAs, such as particle swarm optimization,genetic algorithm, and biogeography-based optimization integrated with ELM, was performed. Based on the outcomes, the ELM-IHHO model exhibits superior performance over the other MOAs in predicting soil compression index.

ATS:A Novel Time-Sharing CPU Scheduling Algorithm Based on Features Similarities

Minimizing time cost in time-shared operating systems is considered basic and essential task,and it is the most significant goal for the researchers who interested in CPU scheduling algorithms.Waiting time,turnaround time,and number of context switches are themost time cost criteria used to compare between CPU scheduling algorithms.CPU scheduling algorithms are divided into non-preemptive and preemptive.RoundRobin(RR)algorithm is the most famous as it is the basis for all the algorithms used in time-sharing.In this paper,the authors proposed a novel CPU scheduling algorithm based on RR.The proposed algorithm is called Adjustable Time Slice(ATS).It reduces the time cost by taking the advantage of the low overhead of RR algorithm.In addition,ATS favors short processes allowing them to run longer time than given to long processes.The specific characteristics of each process are;its CPU execution time,weight,time slice,and number of context switches.ATS clusters the processes in groups depending on these characteristics.The traditionalRRassigns fixed time slice for each process.On the other hand,dynamic variants of RR assign time slice for each process differs from other processes.The essential difference between ATS and the other methods is that it gives a set of processes a specific time based on their similarities within the same cluster.The authors compared between ATS with five popular scheduling algorithms on nine datasets of processes.The datasets used in the comparison vary in their features.The evaluation was measured in term of time cost and the experiments showed that the proposed algorithm reduces the time cost.

Safest Route Detection via Danger Index Calculation and K-Means Clustering

The study aims to formulate a solution for identifying the safest route between any two inputted Geographical locations.Using the New York City dataset,which provides us with location tagged crime statistics;we are implementing different clustering algorithms and analysed the results comparatively to discover the best-suited one.The results unveil the fact that the K-Means algorithm best suits for our needs and delivered the best results.Moreover,a comparative analysis has been performed among various clustering techniques to obtain best results.we compared all the achieved results and using the conclusions we have developed a user-friendly application to provide safe route to users.The successful implementation would hopefully aid us to curb the ever-increasing crime rates;as it aims to provide the user with a beforehand knowledge of the route they are about to take.A warning that the path is marked high on danger index would convey the basic hint for the user to decide which path to prefer.Thus,addressing a social problem which needs to be eradicated from our modern era.

A Novel Binary Emperor Penguin Optimizer for Feature Selection Tasks

Nowadays,due to the increase in information resources,the number of parameters and complexity of feature vectors increases.Optimizationmethods offermore practical solutions instead of exact solutions for the solution of this problem.The Emperor PenguinOptimizer(EPO)is one of the highest performing meta-heuristic algorithms of recent times that imposed the gathering behavior of emperor penguins.It shows the superiority of its performance over a wide range of optimization problems thanks to its equal chance to each penguin and its fast convergence features.Although traditional EPO overcomes the optimization problems in continuous search space,many problems today shift to the binary search space.Therefore,in this study,using the power of traditional EPO,binary EPO(BEPO)is presented for the effective solution of binary-nature problems.BEPO algorithm uses binary search space instead of searching solutions like conventional EPO algorithm in continuous search space.For this purpose,the sigmoidal functions are preferred in determining the emperor positions.In addition,the boundaries of the search space remain constant by choosing binary operators.BEPO’s performance is evaluated over twenty-nine benchmarking functions.Statistical evaluations are made to reveal the superiority of the BEPO algorithm.In addition,the performance of the BEPO algorithm was evaluated for the binary feature selection problem.The experimental results reveal that the BEPO algorithm outperforms the existing binary meta-heuristic algorithms in both tasks.

HARTIV:Human Activity Recognition Using Temporal Information in Videos

Nowadays,the most challenging and important problem of computer vision is to detect human activities and recognize the same with temporal information from video data.The video datasets are generated using cameras available in various devices that can be in a static or dynamic position and are referred to as untrimmed videos.Smarter monitoring is a historical necessity in which commonly occurring,regular,and out-of-the-ordinary activities can be automatically identified using intelligence systems and computer vision technology.In a long video,human activity may be present anywhere in the video.There can be a single ormultiple human activities present in such videos.This paper presents a deep learning-based methodology to identify the locally present human activities in the video sequences captured by a single wide-view camera in a sports environment.The recognition process is split into four parts:firstly,the video is divided into different set of frames,then the human body part in a sequence of frames is identified,next process is to identify the human activity using a convolutional neural network and finally the time information of the observed postures for each activity is determined with the help of a deep learning algorithm.The proposed approach has been tested on two different sports datasets including ActivityNet and THUMOS.Three sports activities like swimming,cricket bowling and high jump have been considered in this paper and classified with the temporal information i.e.,the start and end time for every activity present in the video.The convolutional neural network and long short-term memory are used for feature extraction of temporal action recognition from video data of sports activity.The outcomes show that the proposed method for activity recognition in the sports domain outperforms the existing methods.

Image Segmentation Based on Block Level and Hybrid Directional Local Extrema

In the recent decade,the digitalization of various tasks has added great flexibility to human lifestyle and has changed daily routine activities of communities.Image segmentation is a key step in digitalization.Segmentation plays a key role in almost all areas of image processing,and various approaches have been proposed for image segmentation.In this paper,a novel approach is proposed for image segmentation using a nonuniform adaptive strategy.Region-based image segmentation along with a directional binary pattern generated a better segmented image.An adaptive mask of 8×8 was circulated over the pixels whose bit value was 1 in the generated directional binary pattern.Segmentation was performed in three phases:first,an image was divided into sub-images or image chunks;next,the image patches were taken as input,and an adaptive threshold was generated;and finally the image chunks were processed separately by convolving the adaptive mask on the image chunks.Gradient and Laplacian of Gaussian algorithms along with directional extrema patterns provided a double check for boundary pixels.The proposed approach was tested on chunks of varying sizes,and after multiple iterations,it was found that a block size of 8×8 performs better than other chunks or block sizes.The accuracy of the segmentation technique was measured in terms of the count of ill regions,which were extracted after the segmentation process.

Design of Energy Efficient WSN Using a Noble SMOWA Algorithm

In this paper,the establishment of efficientWireless Sensor Network(WSN)networks has been projected to minimize the consumption of energy using a new Self-adaptive Multi-Objective Weighted Approach(SMOWA)algorithm for solving a multi-objective problem.The Different WSN nodes deployment policies have been proposed and applied in this paper to design an efficientWireless Sensor Network to minimize energy consumption.After that,the cluster head for each cluster has been selected with the help of the duty cycle.After configuring the WSN networks,the SMOWA algorithms have been developed to obtain the minimum energy consumption for the networks.Energy minimization,as well as the amount of day-saving,has been calculated for the differentWSNswhich has been configured through different deployment policies.The major finding of the research paper is to improve the durability of Wireless Sensor Network(i)applying different deployment strategies:(Random,S pattern and nautilus shell pattern),and(ii)using a new Meta-heuristic algorithm(SMOWA Algorithm).In this research,the lifetime of WSN has been increased to a significant level.To choose the best result set from all the obtained results set some constraints such as“equivalent distribution”,“number of repetitions”,“maximum amount energy storage by a node”has been set to an allowable range.

Unleashing the Potential of Civilizational Diplomacy

The China-proposed Global Civilization Initiative has introduced a novel cultural-coexistence framework transcending notions of conflict like the"clash of civilizations"that too often shapes Western diplomacy.