ViBand - High-Fidelity Bio-Acoustic Sensing Using Commodity Smartwatch Accelerometers
Innovative smartwatch technology enhances bio-acoustic sensing, enabling gesture recognition, object detection, and body-based data transmission, expanding wearable interaction possibilities.
Interacting with Soli - Exploring Fine-Grained Dynamic Gesture Recognition in the Radio-Frequency Spectrum
Explore novel machine learning for RF-based gesture recognition using Google's Soli sensor. Learn about high-frequency radar sensing and advanced neural network techniques for fine-grained motion detection.
Mobile Espionage in the Wild - Pegasus and Nation-State Level Attacks
Explore Pegasus, a sophisticated mobile espionage tool, its technical capabilities, and real-world impact. Learn about infrastructure tracking and malware detection in high-profile attacks.
Tracking Ransomware End to End
Explore the rise of ransomware, its revenue model, and tracking methods. Learn about Bitcoin usage, victim payments, and evolving cybercrime tactics in this comprehensive analysis.
Hiding Objects from Computer Vision by Exploiting Correlation Biases
Exploiting correlation biases in computer vision systems to create adversarial images that hide objects from detection, using examples like dogs and frisbees vs. stop signs and pizza.
Detecting Fake 4G Base Stations in Real Time
Learn to detect and understand 4G/LTE IMSI-catchers, explore their capabilities, and discover software solutions for identifying fake base stations. Gain insights into cell network security improvements.
Stegosploit - Drive by Browser Exploits Using Only Images
Innovative technique encoding browser exploits in images, combining steganography and polyglots for undetectable delivery. Explores implementation, challenges, and implications for cybersecurity.
How to Recruit Spies on the Internet
Explore techniques for identifying and recruiting potential spies through online platforms, presented by cybersecurity expert Ira Winkler at LASCON 2012.
A New Philosophy on Artificial Intelligence
Explore a fresh perspective on AI with Kristian Hammond, delving into machine-generated content, context-driven systems, and the future of artificial intelligence in this thought-provoking TEDx talk.
The Accelerating Power of Technology
Explore Ray Kurzweil's vision of future technology, including reverse-engineered human brains and consciousness-operating nanobots, in this thought-provoking talk on technological acceleration.
What Happens When Our Computers Get Smarter Than We Are
Nick Bostrom explores the potential consequences of artificial intelligence surpassing human intelligence, discussing the implications for humanity's future and the importance of aligning AI with human values.
How We Teach Computers to Understand Pictures
Explore the cutting-edge of computer vision as Fei-Fei Li discusses teaching machines to interpret images, showcasing a 15-million-photo database and future possibilities in artificial intelligence.
SkillDetective - Automated Policy-Violation Detection of Voice Assistant Applications in the Wild
Automated tool for detecting policy violations in voice assistant apps, analyzing 54,055 Alexa skills and 5,583 Google actions, identifying over 6,000 violations across 52 policy requirements.
BlockFlex - Enabling Storage Harvesting with Software-Defined Flash in Modern Cloud Platforms
Explore BlockFlex, a learning-based framework for efficient storage harvesting in cloud platforms, enabling fine-grained resource utilization while ensuring performance and security isolation.
Re-architecting Traffic Analysis with Neural Network Interface Cards
Innovative approach to enhance network traffic analysis using binary neural networks on SmartNICs, offering improved scalability, lower latency, and higher throughput compared to traditional software-based systems.
Fine-Tuning Giant Neural Networks on Commodity Hardware with Automatic Pipeline Model Parallelism
Explore fine-tuning giant neural networks on commodity hardware using FTPipe, a system enabling multi-GPU execution through innovative model partitioning and task allocation techniques.
Towards Evaluating the Robustness of Neural Networks
Explores new attack algorithms to assess and challenge the robustness of neural networks against adversarial examples, highlighting vulnerabilities in defensive distillation techniques.
CSI NN - Reverse Engineering of Neural Network Architectures Through Electromagnetic Side Channel
Reverse engineering neural networks through electromagnetic side channels, revealing architecture details like activation functions, layers, neurons, and weights using non-invasive techniques on ARM Cortex-M3 microcontrollers.
Polisis - Automated Analysis and Presentation of Privacy Policies Using Deep Learning
Automated framework for analyzing privacy policies using deep learning, enabling scalable queries and applications like privacy icon assignment and question-answering systems.
Distillation as a Defense to Adversarial Perturbations against Deep Neural Networks
Introduces defensive distillation to protect deep neural networks from adversarial attacks, reducing attack effectiveness from 95% to 0.5% and increasing robustness by 800% in tests.
Model Orthogonalization - Class Distance Hardening in Neural Networks for Better Security
Explore a novel technique to enhance neural network security through class distance hardening, improving model robustness against adversarial attacks and ensuring more reliable AI systems.