Search for: All records

Research in the upcoming field of adversarial ML has revealed that machine learning, especially deep learning, is highly vulnerable to imperceptible adversarial perturbations, both in the domain of vision as well as speech. This has induced an urgent need to devise fast and practical approaches to secure deep learning models from adversarial attacks, so that they can be safely deployed in real-world applications. In this showcase, we put forth the idea of compression as a viable solution to defend against adversarial attacks across modalities. Since most of these attacks depend on the gradient of the model to craft an adversarial instance, compression, which is usually non-differentiable, denies a useful gradient to the attacker. In the vision domain we have JPEG compression, and in the audio domain we have MP3 compression and AMR encoding -- all widely adopted techniques that have very fast implementations on most platforms, and can be feasibly leveraged as defenses. We will show the effectiveness of these techniques against adversarial attacks through live demonstrations, both for vision as well as speech. These demonstrations would include real-time computation of adversarial perturbations for images and audio, as well as interactive application of compression for defense. We would invite and encourage the audience to experiment with their own images and audio samples during the demonstrations. This work was undertaken jointly by researchers from Georgia Institute of Technology and Intel Corporation. 

Abstract Clean and efficient energy storage and conversion via sustainable water and nitrogen reactions have attracted substantial attention to address the energy and environmental issues due to the overwhelming use of fossil fuels. These electrochemical reactions are crucial for desirable clean energy technologies, including advanced water electrolyzers, hydrogen fuel cells, and ammonia electrosynthesis and utilization. Their sluggish reaction kinetics lead to inefficient energy conversion. Innovative electrocatalysis, i.e., catalysis at the interface between the electrode and electrolyte to facilitate charge transfer and mass transport, plays a vital role in boosting energy conversion efficiency and providing sufficient performance and durability for these energy technologies. Herein, a comprehensive review on recent progress, achievements, and remaining challenges for these electrocatalysis processes related to water (i.e., oxygen evolution reaction, OER, and oxygen reduction reaction, ORR) and nitrogen (i.e., nitrogen reduction reaction, NRR, for ammonia synthesis and ammonia oxidation reaction, AOR, for energy utilization) is provided. Catalysts, electrolytes, and interfaces between the two within electrodes for these electrocatalysis processes are discussed. The primary emphasis is device performance of OER‐related proton exchange membrane (PEM) electrolyzers, ORR‐related PEM fuel cells, NRR‐driven ammonia electrosynthesis from water and nitrogen, and AOR‐related direct ammonia fuel cells.