Search by Subject - Artificial Intelligence, Machine Learning, Computer Vision, Natural language processing (2024)

Building Detection-Resistant Reconnaissance Attacks Based on Adversarial Explainability

• Mohammed M. Alani,
• Atefeh Mashatan,
• Ali Miri

The growing popularity of Internet-of-Things devices makes them a desired target for malicious actors. Most attacks start with a reconnaissance phase where the attacker gathers information about the services running on the device, the open ports, and any ...

Improving Steganographic Security with Source Biasing

• Eli Dworetzky,
• Edgar Kaziakhmedov,
• Jessica Fridrich

By selecting covers in which steganographic embedding is harder to detect, the steganographer can decrease the chances of being caught by the Warden. On the other hand, sampling from the cover source with a bias is detectable on its own. In this paper, ...

Exploring Diffusion-Inspired Pixel Predictors for WS Steganalysis

• Martin Beneš,
• Rainer Böhme

Analytical estimators of the steganographic change rate in images, such as WS steganalysis, often operate on the noise residual. The residual can be obtained by estimating the cover content with pixel predictors and subtracting it from the image under ...

Investigating Translation Invariance and Shiftability in CNNs for Robust Multimedia Forensics: A JPEG Case Study

• Edoardo Daniele Cannas,
• Sara Mandelli,
• Paolo Bestagini,
• Stefano Tubaro

Convolutional Neural Networks (CNNs) have been the state of the art in many applications, including computer vision and multimedia forensics. Translation invariance is often included among the reasons for their success. However, the recent literature has ...

Suppressing High-Frequency Artifacts for Generative Model Watermarking by Anti-Aliasing

• Li Zhang,
• Yong Liu,
• Xinpeng Zhang,
• Hanzhou Wu

Protecting deep neural networks (DNNs) against intellectual property (IP) infringement has attracted an increasing attention in recent years. Recent advances focus on IP protection of generative models, which embed the watermark information into the ...

WikiPhish: A Diverse Wikipedia-Based Dataset for Phishing Website Detection: Data/Toolset Paper

• Gabriel Loiseau,
• Valentin Lefils,
• Maxime Meyer,
• Damien Riquet

Phishing remains a pervasive security threat, necessitating effective and universally comparable detection systems. The use of supervised machine learning models for phishing detection has been generalized in the literature to automate predictions and ...

SAMANTHA: A chatbot to assist users in training tasks to prevent workplace hazards

• David Contreras Aguilar,
• Fernando Medina,
• Mauricio Oyanedel,
• Maria Salamó,
• Miquel Sànchez-Marrè

In businesses, preventing workplace hazards becomes crucial. In order to limit negative effects on people, society, and the economy, it is crucial for both the organization and its employees to reduce accidents and occupational illnesses. Staff training ...

Insight AI Risk Detection Model - Vulnerable People Emotional Situation Support

• Diego Gosmar,
• Elena Peretto,
• Oita Coleman

This paper presents an AI-based risk detection model (architectural framework) for real-time emotional support and risk assessment, addressing the rise in mental health issues among youth. The model leverages Insight AI (Sentiment and Emotional Analysis) ...

Leveraging Large Language Models for Preliminary Security Risk Analysis: A Mission-Critical Case Study

• Matteo Esposito,
• Francesco Palagiano

Preliminary security risk analysis (PSRA) provides a quick approach to identify, evaluate, and propose remediation to potential risks in specific scenarios. The extensive expertise required for an effective PSRA and the substantial textual-related tasks ...

Trustworthy AI in practice: an analysis of practitioners' needs and challenges

• Maria Teresa Baldassarre,
• Domenico Gigante,
• Marcos Kalinowski,
• Azzurra Ragone,
• Sara Tibidò

Recently, there has been growing attention on behalf of both academic and practice communities towards the ability of Artificial Intelligence (AI) systems to operate responsibly and ethically. As a result, a plethora of frameworks and guidelines have ...

AI-enabled efficient PVM performance monitoring

• Mario Veniero,
• Davide Varriale

The recent expansion of photovoltaic (PV) systems and increased production scale necessitate enhanced monitoring to assess system performance, detect potential degradation, and identify imminent failures, ensuring sustained quality and optimal ...

Motivation Research Using Labeling Functions

• Idan Amit,
• Dror G. Feitelson

Motivation is an important factor in software development. However, it is a subjective concept that is hard to quantify and study empirically. In order to use the wealth of data available about real software development projects in GitHub, we represent ...

Assessing healthcare software built using IoT and LLM technologies

• Gabriele De Vito

In the fast-paced world of healthcare technology, combining IoT devices with large language models (LLMs) offers a promising path to transform Clinical Decision-Support Systems (CDSS). This Ph.D. project is designed to tap into IoT’s extensive data ...

Why Large Language Models will (not) Kill Software Engineering Research

• Massimiliano Di Penta

Over the last decade, we have witnessed a flourishing activity in the application of deep learning techniques to solve software engineering problems that were poorly addressed in the past, or not addressed at all. In this context, researchers put effort ...

Detecting Security Fixes in Open-Source Repositories using Static Code Analyzers

• Therese Fehrer,
• Rocio Cabrera Lozoya,
• Antonino Sabetta,
• Dario Di Nucci,
• Damian A. Tamburri

Surfing the AI Wave in Software Engineering: Opportunities and Challenges

• Nicole Novielli

The diffusion of generative AI, specifically Large Language Models (LLMs), is profoundly affecting Software Engineering. Thanks to their unprecedented potential for disruptive changes, which mainly reside in their ability to reduce the need for large-...

System Optimizations for Enabling Training of Extreme Long Sequence Transformer Models

• Sam Ade Jacobs,
• Masahiro Tanaka,
• Chengming Zhang,
• Minjia Zhang,
• Reza Yazdani Aminadabi,
• Shuaiwen Leon Song,
• Samyam Rajbhandari,
• Yuxiong He

Computation in a typical Transformer-based large language model (LLM) can be characterized by batch size, hidden dimension, number of layers, and sequence length. Until now, system works for accelerating LLM training have focused on the first three ...

ExTra CTI: Explainable and Transparent Child-Technology Interaction

• Elmira Yadollahi,
• Mike E.U. Ligthart,
• Ksh*tij Sharma,
• Elisa Rubegni

When the technology encompasses some form of intelligence or agency in the form of robots, virtual agents or artificial intelligence, understanding the reasoning behind their actions and decisions becomes an integral part of the interaction. This ...

Toward Personalised Learning Experiences: Beyond Prompt Engineering

• Joost Kruis,
• Maria Soledad Pera,
• Zoë ten Napel,
• Monica Landoni,
• Emiliana Murgia,
• Theo Huibers,
• Remco Feskens

We discuss the foundation of a collaborative effort to explore AI’s role in supporting (teachers and) children in their learning experiences. We integrate principles of educational psychology, AI, and HCI, and align with best practices in education ...

Batch Active Learning of Reward Functions from Human Preferences

• Erdem Biyik,
• Nima Anari,
• Dorsa Sadigh

Data generation and labeling are often expensive in robot learning. Preference-based learning is a concept that enables reliable labeling by querying users with preference questions. Active querying methods are commonly employed in preference-based ...