Securing AI: Understanding and Defending Against Adversarial Attacks in Deep Learning Systems
Abstract
This review paper delves into the intricate landscape of security vulnerabilities within deep learning frameworks, specifically focusing on adversarial attacks and their impact across diverse AI applications. It scrutinizes vulnerabilities in neural network models, reinforcement learning policies, Natural Language Processing (NLP) classifiers, cloud-based image detectors, and deep convolutional neural networks (CNNs). The paper illuminates’ techniques such as adversarial example generation and their applicability in exploiting vulnerabilities in various scenarios, underlining the imperative need for robust defense mechanisms. Additionally, it explores innovative methodologies like influence functions and outlier detection to enhance understanding, debug models, and fortify defenses
against adversarial attacks. The paper concludes by accentuating the critical importance of addressing these vulnerabilities and fostering further research in securing AI systems against potential threats. Absolutely! Here a simpler abstract that captures the essence of your review paper: It looks at how sneaky tricks can fool smart AI systems. It talks about how bad guys can make AI mess up, even in important things like self-driving cars, language understanding, and image recognition. The paper shows different ways these tricks work and how they can be used against various types of AI. It also shares some cool ideas to make AI safer and tougher against these tricks. The paper ends by saying it really important to make AI safer from these sneaky attacks.
Keywords:
Deep Learning Security, Vulnerabilities in AI Systems, Neural Network Vulnerability, Reinforcement Learning Vulnerabilities, Adversarial Examples, Defense Mechanisms in Deep Learning, Natural Language Processing (NLP) Security, Cloud-Based Image Detectors, Convolutional Neural Networks (CNNs) Vulnerabilities, Machine Learning Security Risks, Adversarial Examples in Physical World, Interpretability of Deep Neural Networks, Obfuscated Gradients, Defense Strategies against Adversarial AttacksPublished
Issue
Section
License
Copyright (c) 2023 International Journal on Emerging Research Areas

This work is licensed under a Creative Commons Attribution 4.0 International License.
All published work in this journal is licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0). This license permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
How to Cite
Similar Articles
- Jefrin Siby Mathew, Joyal Joseph, Roshik George, Tinu Rose Thottungal , Honey Joseph, Multiple Disease Detection using Machine Learning , International Journal on Emerging Research Areas: Vol. 3 No. 1 (2023): IJERA
- Dr. Indu John, Gauri Santhosh, Jesna Susan Reji, Abdul Musawir, Glady Prince, Detection of Autism Spectrum Disorder in Toddlers using Machine Learning , International Journal on Emerging Research Areas: Vol. 4 No. 1 (2024): IJERA
- Nighila Ashok, Adithya Ajith, Aparna Shaju, Arjuna Chandran V V, Fahmi Fathima T S, DeepScan : A Deepfake Video Detection System , International Journal on Emerging Research Areas: Vol. 4 No. 1 (2024): IJERA
- Akhil Mathew Mohan, Alan Maria George, Arathy Baby, Gopika S, Syam Gopi, Abubeker K.M, Real-time Air Quality Index Monitoring and Alert System using IoT Technology , International Journal on Emerging Research Areas: Vol. 3 No. 1 (2023): IJERA
- Amala Jayan, Feneesha V B, Rameesa Dilsa C P, Sandra Maryam Binu, Sandra Maryam Binu, Stockwise: A survey on stock price prediction models , International Journal on Emerging Research Areas: Vol. 4 No. 1 (2024): IJERA
- FATHIMA P.S, ANU ROSE JOY, ANSPIN TITUS, ANSU MARIUM SHIBU, ASNA AZEEZ, INDIAN SIGN LANGUAGE RECOGNITION USING YOLOV5 , International Journal on Emerging Research Areas: Vol. 4 No. 1 (2024): IJERA
- Kaveri S, Pooja Satheesh, Kesiya Susan John, Reubel K Wilson, Dr. Jacob John, Predictive Maintenance of Machines Using IoT and Machine Learning , International Journal on Emerging Research Areas: Vol. 4 No. 2 (2024): IJERA
- Adams Mathew, Adithya Sanil, Akhil J Medackal, Nikhil J Medackal, Dyni Thomas, A Literature Review on IMAGE FORGERY DETECTION , International Journal on Emerging Research Areas: Vol. 3 No. 1 (2023): IJERA
- Layana S Pradeep, Milen Ninan Ittiyeipe, Shahina S, Soumya A S, Ojus Thomas Lee , Gayathri Mohan, A REVIEW OF LOAD ESTIMATION AND DISTRIBUTION STRATEGY FOR RENEWABLE ENERGY SOURCES , International Journal on Emerging Research Areas: Vol. 4 No. 1 (2024): IJERA
- Joel Lee George, Karthik S Kumar , Riya Merce Thomas, Roshan Roy Varghese, Simy Mary Kurian, Epidemo A Machine Learning Regression-Based , International Journal on Emerging Research Areas: Vol. 4 No. 1 (2024): IJERA
You may also start an advanced similarity search for this article.