The Research Forum showcased ‘work-in-progress’ and short research talks on emerging hot topics, preliminary research results, practical problems encountered by industry practitioners, lessons learned, exciting datasets, and other topics of interest to security researchers working in deep learning.
Seven authoring teams received Outstanding Research Talk awards with voucher prizes and a top prize of a Black Hat Asia 2018 Briefings Pass. Find out the winning teams here!
"User Action as An Authentication Mechanism to Improve Smartphone Security" (Presentation slides)
Arul Prakash Samathuvamani
"Privacy Leakage In Long Short Term Memory" (Presentation slides)
"Attacking the Image Captioning Model - Take Show and Tell Model As An Example" (Presentation slides)
"Comparative Evaluation Of Synthetic Data Generation Methods" (Presentation slides)
Ashish Dandekar, Remmy A. M. Zen, and Stephane Bressan
"Side-Channel Analysis and Machine Learning: A Practical Perspective" (Presentation slides)
Sylvain Guilley and Matthieu Lec'hvien
"On the Search for Invertible Generative Adversarial Networks"
Jiyi Zhang, Hung Dang, Hwee Kuan Lee, and Ee-Chien Chang
"On the Application of Deep Learning Techniques to Website Fingerprinting Attacks and Defenses" (Presentation slides)
Marc Juarez and Vera Rimmer
"Recovering Types From Binaries" (Presentation slides)
Teodora Baluta, Shiqi Shen, and Alexandros Dimos
"SecHadoop: End-to-End Privacy Preserving Hadoop" (Presentation slides)
"Learning Relations Between Variables Using Deep Learning"
Shiqi Shen and Soundarya Ramesh
"A Try at Task-Based Dialogue Building" (Presentation slides)
"Early Detection of Crossfire Attacks using Deep Learning" (Presentation slides)
Saurabh Misra, Mengxuan Tan, Mostafa Rezazad, and Ngai-Man Cheung
"Do We Need Original Data for Training? Toward Designing Privacy-Preserving Machine Learning" (Presentation slides)
Qingrong Chen, Minhui Xue, Chong Xiang, Bo Li, Haizhong Zheng, and Haojin Zhu
"Neural Architecture Search: Insights and Long-Term Horizons" (Presentation slides)
"Noise Data Augmentation for Speaker Recognition using Conditional Generative Adversarial Networks" (Presentation slides)
"On Lyapunov Exponents and Adversarial Perturbations" (Presentation slides)
Vinay Prabhu and John Whaley
"Transferability of Adversarial Attacks in Model-Agnostic Meta-Learning" (Presentation slides)
Riley Edmunds, Noah Golmant, Vinay Ramasesh, Phillip Kuznetsov, Piyush Patil, and Raul Puri
SESSION I: DEEP LEARNING FOR SECURITY
"An Implementation of Web Application Firewall Based on a Deep Neural Network Detection Engine"
Liang Shi, Staff Expert and Manager of Security Data Science team, Alibaba Cloud Security
Min Ye, Senior Security Expert, Alibaba Cloud Security
Tianlong Liu, Senior Algorithm Engineer, Alibaba Cloud Security
"Deep Learning for User Authentication " (Presentation slides)
Dr John Whaley, Founder and CEO, UnifyID
SESSION II: SECURITY FOR DEEP LEARNING
"Data Privacy in Machine Learning" (Presentation slides)
Dr Reza Shokri, Assistant Professor, National University of Singapore
"Adversarial Deep Learning: Attacks and Defenses"
Dr Ian Fischer, Researcher, Google Research
"Demo on Biometrics Security" (Presentation slides)
Gao Shupeng, Baidu