Deep Learning - Beyond the Hype
Explore neural networks and reinforcement learning for game AI development, with insights on their potential and practical applications in the gaming industry.
Robocalypse Now - Using Deep Learning to Combat Cheating in Counter-Strike - Global Offensive
Explore Valve's innovative use of Deep Learning to detect and prevent cheating in Counter-Strike: Global Offensive, revolutionizing game integrity and fair play.
Lip-Interact - Improving Mobile Device Interaction with Silent Speech Commands
Innovative technique using silent speech commands for smartphone control, enabling efficient one-handed input and enhancing privacy through front camera lip movement recognition.
Deep Learning with Differential Privacy
Explore deep learning with differential privacy, focusing on techniques to protect data privacy while maintaining model accuracy in machine learning applications.
Playing Offense and Defense with Deepfakes
Explore deepfake technology's offensive and defensive applications, including creation process and underlying deep learning models. Gain insights into this powerful and controversial AI tool.
Alpa - Automating Inter- and Intra-Operator Parallelism for Distributed Deep Learning
Automating model-parallel training for large deep learning models through a hierarchical approach, unifying data, operator, and pipeline parallelism for efficient distributed execution.
Accelerating Collective Communication in Data Parallel Training across Deep Learning Frameworks
Techniques to improve Horovod's communication efficiency in data parallel training, enhancing performance and scalability for deep learning frameworks on supercomputers.
SHADE - Enable Fundamental Cacheability for Distributed Deep Learning Training
Innovative caching system for distributed deep learning training, improving performance by prioritizing important data samples and enhancing cache hit ratios in I/O-intensive workloads.
Rammer - Enabling Holistic Deep Learning Compiler Optimizations with rTasks
Optimizing DNN computation on hardware accelerators through novel abstractions and holistic scheduling, achieving significant performance improvements over existing compilers and libraries.
Making and Measuring Progress in Adversarial Machine Learning
Explore adversarial machine learning progress with Nicholas Carlini, focusing on measurement techniques and recent advancements in this critical field of AI security research.
Targeted Adversarial Examples for Black Box Audio Systems
Exploring techniques to generate adversarial examples for audio systems, combining genetic algorithms and gradient estimation to achieve high attack success rates while maintaining audio similarity.
Efficient Evaluation of Activation Functions over Encrypted Data
Explore techniques for approximating activation functions on encrypted data, with applications in secure machine learning for image classification and representation learning.
Deep in the Dark - Deep Learning-based Malware Traffic Detection without Expert Knowledge
Explores deep learning models for malware traffic detection using raw network data, comparing performance to traditional approaches and evaluating different input representations.
Antennen vermessen
Praktische Einführung in die Vermessung und Optimierung von Antennen für verschiedene Funktechnologien. Erklärt Unterschiede zwischen Antennentypen, Kabeln und Filtern für bessere Leistung.
Graph Neural Networks for Information Extraction with PyTorch
Explore graph neural networks for information extraction using PyTorch, focusing on applications in NLP and understanding tabular documents through spatial encoding of word relationships.
Deepfakes - If Anything Can Be Real, Then Nothing Is Real
Explore the implications of deepfake technology on reality perception, truth, and trust in media. Learn about detection methods and the future of synthetic media.
Garbage In, Garbage Out - How Purportedly Great Machine Learning Models Can Be Screwed Up by Bad Data
Explores how data quality impacts machine learning models for malicious URL detection, comparing results across different datasets and analyzing feature activations in neural networks.
Ensuring a Thriving Digital Future in a Post-Truth World
Futurist Mark van Rijmenam explores AI's impact on truth, presenting strategies for navigating deepfakes and language models. He offers a three-pronged approach to shape a thriving digital future in our post-truth era.
Composing Bach Chorales Using Deep Learning
Explore how deep learning and LSTM can compose Bach-style chorales, with insights from a musical Turing test revealing AI's potential for musical creativity.
Breaking Black-box AI
Explore failure cases of popular machine learning algorithms through live demos, from linear regression to deep learning, revealing limitations and potential pitfalls in AI applications.
Unleashing the Power of Decompilation for x86 Deep Neural Network Executables
Explore BTD, a decompiler for deep neural network executables, enabling reverse engineering of compiled DL models on various hardware platforms to enhance security and transparency.