Results 61 to 70 of about 4,795 (257)

Exploring Synergy of Denoising and Distillation: Novel Method for Efficient Adversarial Defense

Applied Sciences
Escalating advancements in artificial intelligence (AI) has prompted significant security concerns, especially with its increasing commercialization. This necessitates research on safety measures to securely utilize AI models.
Inpyo Hong, Sokjoon Lee
doaj   +1 more source

Legal Professionals Use an Implicit Checklist to Assess Expert Witness Credibility

Behavioral Sciences &the Law, EarlyView.
ABSTRACT In a court of law, some forms of evidence are too technical to interpret without the help of an expert witness. However, expert testimony sometimes leads to wrongful conviction. We used an online survey to quantitatively compare how legal professionals (n = 122) and lay people (n = 109) understood what makes an expert witness credible.
Mahensingh Deonaran, Yvan I. Russell, Mandeep K. Dhami   +2 more
wiley   +1 more source

Adversarial Backdoor Defense in CLIP

CoRR
Multimodal contrastive pretraining, exemplified by models like CLIP, has been found to be vulnerable to backdoor attacks. While current backdoor defense methods primarily employ conventional data augmentation to create augmented samples aimed at feature alignment, these methods fail to capture the distinct features of backdoor samples, resulting in ...
Junhao Kuang, Siyuan Liang 0004, Jiawei Liang, Kuanrong Liu, Xiaochun Cao   +4 more
openaire   +2 more sources

Generative Adversarial Trainer: Defense to Adversarial Perturbations with GAN

CoRR, 2017
We propose a novel technique to make neural network robust to adversarial examples using a generative adversarial network. We alternately train both classifier and generator networks. The generator network generates an adversarial perturbation that can easily fool the classifier network by using a gradient of each image.
Hyeungill Lee, Sungyeob Han, Jungwoo Lee 0001   +2 more
openaire   +2 more sources

The Role of Trade Unions in Environmental Governance: Exploring Workers Legitimacy Judgments

Environmental Policy and Governance, EarlyView.
ABSTRACT Trade unions are increasingly expected to play a role in sustainability transitions, but their legitimacy as environmental governance actors remains underexplored. Our study uses a multidimensional framework to examine how workers, both union members and non‐members, evaluate trade union legitimacy “from below.” This framework distinguishes ...
Urša Golob, Branko Bembič, Klement Podnar   +2 more
wiley   +1 more source

A divide-and-conquer reconstruction method for defending against adversarial example attacks

Visual Intelligence
In recent years, defending against adversarial examples has gained significant importance, leading to a growing body of research in this area. Among these studies, pre-processing defense approaches have emerged as a prominent research direction. However,
Xiyao Liu, Jiaxin Hu, Qingying Yang, Ming Jiang, Jianbiao He, Hui Fang   +5 more
doaj   +1 more source

Adversarial Attacks and Defense on Texts: A Survey

CoRR, 2020
Deep learning models have been used widely for various purposes in recent years in object recognition, self-driving cars, face recognition, speech recognition, sentiment analysis, and many others. However, in recent years it has been shown that these models possess weakness to noises which force the model to misclassify.
Aminul Huq, Mst. Tasnim Pervin
openaire   +2 more sources

Beyond Prediction: Data, Baselines, Explanation, and Causation in Machine Learning for Food Insecurity

Food Safety and Health, EarlyView.
The gains from machine learning in nowcasting and forecasting food insecurity are still small and limited and cannot be observed in all countries. This review summarizes the public resources for data, corrects common misconceptions about model requirements, and establishes baseline requirements, explanations, causal inference, and equity in operational
Shabnam Mehboob, Azman Abdullah, Gunesh Kumar, Samreen Memon, Naila Afghan   +4 more
wiley   +1 more source

An enhanced ensemble defense framework for boosting adversarial robustness of intrusion detection systems

Scientific Reports
Machine learning (ML) and deep neural networks (DNN) have emerged as powerful tools for enhancing intrusion detection systems (IDS) in cybersecurity.
Zeinab Awad, Magdy Zakaria, Rasha Hassan
doaj   +1 more source

Citrus Essential Oils in Food Safety and Health: Advances in AI‐Assisted Authentication, Antimicrobial Applications, Sustainable Extraction, and Regulatory Translation

Food Safety and Health, EarlyView.
Citrus essential oils combine sustainable extraction, AI‐assisted authentication, multi‐omics profiling, antimicrobial activity, and regulatory safety to enhance food quality, detect adulteration, and promote consumer health, and support intelligent, sustainable food systems through innovative industrial applications and circular bioeconomy strategies.
Md. Hassan Bin Nabi, Liza Akter, Yusuf Chowdhury, Mohima Akther Any, Nazmul Islam, Erni Sofia Murtini, Nurul Huda, Wahidu Zzaman   +7 more
wiley   +1 more source