Adversarial Attacks on Face Detectors using Neural Net based Constrained Optimization
|
|
|
- Lora Lindsey
- 5 years ago
- Views:
Transcription
1 Adversarial Attacks on Face Detectors using Neural Net based Constrained Optimization Joey Bose University of Toronto September 26, 2018 Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
2 Motivation Machine Learning models are Ubiquitous Generalization behavior of Deep Neural Nets is still very poorly understood Attacking models reveals weaknesses and drives research towards Robust Models Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
3 Attacking the Machine Learning Pipeline Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
4 Adversarial Attacks - Basic Phenomena minimize L(x, x + δ) s.t. D(x + δ) = t x + δ [0, 1] n Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
5 Early Attacks - FGSM (Goodfellow et al. 2014) Given an image x, the Fast Gradient Sign Method (FGSM) returns a perturbed input x : x = x ɛ sign( x J(θ, x, y)) where J is the loss function for the attacked classifier and ɛ controls the extent of the perturbation. Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
6 FGSM on MNIST Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
7 Basic Iterative Method on MNIST (Kurakin et. al 2016) Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
8 Carlini-Wagner (Carlini and Wagner 2016) Find some small δ such that D(x + δ) = t, argmin δ δ p + c f (x + δ) s.t. x + δ [0, 1] n where f is an objective function such that D(x + δ) = t f (x + δ) 0. The Carlini-Wagner attack is very strong achieving over 99.8% misclassification on CIFAR-10 but is slow and computationally expensive Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
9 Adversarial Transformation Networks (Baluja et al. 2017) Adversarial Transformative Network (ATN) is any neural network that, given an input image, returns an adversarial image: argmin θ β L X (g f,θ (x i ), x i ) + L Y (f (g f,θ (x i )), f (x i )) x i X where β is a scalar, L X is a perceptual loss (e.g., the L 2 distance) between the original and perturbed inputs and L Y is the loss between the classifier s predictions on the original inputs and the perturbed inputs. ATNs were less effective than strong attacks like Carlini-Wagner ATN s are fast, adversarial image can be created with just a forward pass through the ATN ATN s adversarial images are not transferable Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
10 Object Detection in Pictures Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
11 Faster RCNN Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
12 Adversarial Attacks on Object Detection Object Detectors are much harder to attack than classification models due to: Number of Targets in an Image are much higher A successful attack must fool ALL Proposed Bounding Boxes Older Detectors are not always end to end differentiable Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
13 Problem Setup Constructing adversarial examples for face detectors can be framed as a constrained optimization problem similar to the Carlini-Wagner attack. minimize L(x, x + δ) s.t. D(x + δ) = t x + δ [ 1, 1] n This optimization problem is typically very difficult as the constraint D(x + δ) = t is highly non-linear due to D being a neural network. Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
14 Relaxation The constraint can be moved to the objective function as a penalty term for violating the original constraint. minimize L(x, x + δ) + λl misclassify (x + δ) s.t. x + δ [ 1, 1] n The constant λ > 0 balances the magnitude of the perturbation generated to the actual adversarial goal. Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
15 Approach Motivation Optimizing over a single parameter per image is still difficult for a detection network. Adversarial attacks against face detectors should perturb pixels mostly on face regions Learning abstract representations of a face should help constructing attacks on new faces Fast generation of adversarial images enables Adversarial Training Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
16 Threat Model Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
17 Choice of Misclassification Loss There are many possible choices for Misclassification Loss Likelihood of perturbed images under D N i=1 max(0, Z(x i ) face Z(x i ) background) N i=1 max(0, D(x i ) face D(x i ) background) Empirically, some loss functions are better than others as the constant λ is either too small or too large during different phases in training. Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
18 Learning the Generator L total (x, x ) = x x λ N max(0, Z(x i ) face Z(x i ) background ) (1) i=1 Conditional generator G is trained using a pretrained detector over ALL targets proposed by the detector. Spending more time on a given example allows greatly stabilizes training Choosing the same misclassification loss as the Carlini Wagner attack is more robust to the choice of λ. Training was not successful otherwise. Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
19 Implementation Details Input are resized to a resolution of 600 by 800 pixels The number of object proposals are restricted to a maximum of 2000 during training and 300 during test Only Object proposals with probability greater than α = 0.7 are considered We pre train our Faster R-CNN face detector on the WIDER face dataset for 14 epochs with the ADAM optimizer Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
20 Attacks on Cropped 300-W Dataset Faster R-CNN Our Attack α = α = α = α = α = α = Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
21 Results Runtime Joey Bose (UofT) FGSM 2.21s C-W >6300s GeekPwn Las Vegas Ours 1.21s September 26, / 27
22 More Results Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
23 More Results Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
24 More Results Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
25 Video Demo Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
26 Attacks under JPEG compression Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
27 Ongoing and Future Research Directions Extend attack to multiple detectors Construct a Black-box variation of this attack using Policy Gradients Characterize the space of adversarial examples between two detectors. Joey Bose (UofT) GeekPwn Las Vegas September 26, / 27
Adversarial examples in Deep Neural Networks. Luiz Gustavo Hafemann Le Thanh Nguyen-Meidine
Adversarial examples in Deep Neural Networks Luiz Gustavo Hafemann Le Thanh Nguyen-Meidine Agenda Introduction Attacks and Defenses NIPS 2017 adversarial attacks competition Demo Discussion 2 Introduction
Lesson 08. Convolutional Neural Network. Ing. Marek Hrúz, Ph.D. Katedra Kybernetiky Fakulta aplikovaných věd Západočeská univerzita v Plzni.
Lesson 08 Convolutional Neural Network Ing. Marek Hrúz, Ph.D. Katedra Kybernetiky Fakulta aplikovaných věd Západočeská univerzita v Plzni Lesson 08 Convolution we will consider 2D convolution the result
Adversarial Examples and Adversarial Training. Ian Goodfellow, OpenAI Research Scientist Presentation at Quora,
Adversarial Examples and Adversarial Training Ian Goodfellow, OpenAI Research Scientist Presentation at Quora, 2016-08-04 In this presentation Intriguing Properties of Neural Networks Szegedy et al, 2013
Adversarial Examples and Adversarial Training. Ian Goodfellow, OpenAI Research Scientist Presentation at HORSE 2016 London,
Adversarial Examples and Adversarial Training Ian Goodfellow, OpenAI Research Scientist Presentation at HORSE 2016 London, 2016-09-19 In this presentation Intriguing Properties of Neural Networks Szegedy
On the Robustness of Deep Neural Networks
On the Robustness of Deep Neural Networks Manuel Günther, Andras Rozsa, and Terrance E. Boult Vision and Security Technology Lab, University of Colorado Colorado Springs {mgunther,arozsa,tboult}@vast.uccs.edu
Defense Against the Dark Arts: Machine Learning Security and Privacy. Ian Goodfellow, Staff Research Scientist, Google Brain BayLearn 2017
Defense Against the Dark Arts: Machine Learning Security and Privacy Ian Goodfellow, Staff Research Scientist, Google Brain BayLearn 2017 An overview of a field This presentation summarizes the work of
arxiv: v1 [cs.cv] 12 Jul 2017
![arxiv: v1 [cs.cv] 12 Jul 2017](/thumbs/79/79524156.jpg "arxiv: v1 [cs.cv] 12 Jul 2017")
NO Need to Worry about Adversarial Examples in Object Detection in Autonomous Vehicles Jiajun Lu, Hussein Sibai, Evan Fabry, David Forsyth University of Illinois at Urbana Champaign {jlu23, sibai2, efabry2,
Supplementary Material for Generative Adversarial Perturbations
Supplementary Material for Generative Adversarial Perturbations Omid Poursaeed 1,2 Isay Katsman 1 Bicheng Gao 3,1 Serge Belongie 1,2 1 Cornell University 2 Cornell Tech 3 Shanghai Jiao Tong University
Lecture 17 Convolutional Neural Networks
Lecture 17 Convolutional Neural Networks 30 March 2016 Taylor B. Arnold Yale Statistics STAT 365/665 1/22 Notes: Problem set 6 is online and due next Friday, April 8th Problem sets 7,8, and 9 will be due
Introduction to Machine Learning
Introduction to Machine Learning Deep Learning Barnabás Póczos Credits Many of the pictures, results, and other materials are taken from: Ruslan Salakhutdinov Joshua Bengio Geoffrey Hinton Yann LeCun 2
An Introduction to Convolutional Neural Networks. Alessandro Giusti Dalle Molle Institute for Artificial Intelligence Lugano, Switzerland
An Introduction to Convolutional Neural Networks Alessandro Giusti Dalle Molle Institute for Artificial Intelligence Lugano, Switzerland Sources & Resources - Andrej Karpathy, CS231n http://cs231n.github.io/convolutional-networks/
Enhancing Symmetry in GAN Generated Fashion Images
Enhancing Symmetry in GAN Generated Fashion Images Vishnu Makkapati 1 and Arun Patro 2 1 Myntra Designs Pvt. Ltd., Bengaluru - 560068, India vishnu.makkapati@myntra.com 2 Department of Electrical Engineering,
Deep Neural Networks (2) Tanh & ReLU layers; Generalisation and Regularisation
Deep Neural Networks (2) Tanh & ReLU layers; Generalisation and Regularisation Steve Renals Machine Learning Practical MLP Lecture 4 9 October 2018 MLP Lecture 4 / 9 October 2018 Deep Neural Networks (2)
Analysis of adversarial attacks against CNN-based image forgery detectors
Analysis of adversarial attacks against CNN-based image forgery detectors Diego Gragnaniello, Francesco Marra, Giovanni Poggi, Luisa Verdoliva Department of Electrical Engineering and Information Technology
Learning Deep Networks from Noisy Labels with Dropout Regularization
Learning Deep Networks from Noisy Labels with Dropout Regularization Ishan Jindal*, Matthew Nokleby*, Xuewen Chen** *Department of Electrical and Computer Engineering **Department of Computer Science Wayne
Autocomplete Sketch Tool
Autocomplete Sketch Tool Sam Seifert, Georgia Institute of Technology Advanced Computer Vision Spring 2016 I. ABSTRACT This work details an application that can be used for sketch auto-completion. Sketch
Augmenting Self-Learning In Chess Through Expert Imitation
Augmenting Self-Learning In Chess Through Expert Imitation Michael Xie Department of Computer Science Stanford University Stanford, CA 94305 xie@cs.stanford.edu Gene Lewis Department of Computer Science
Deep Learning. Dr. Johan Hagelbäck.
Deep Learning Dr. Johan Hagelbäck johan.hagelback@lnu.se http://aiguy.org Image Classification Image classification can be a difficult task Some of the challenges we have to face are: Viewpoint variation:
Training a Minesweeper Solver
Training a Minesweeper Solver Luis Gardea, Griffin Koontz, Ryan Silva CS 229, Autumn 25 Abstract Minesweeper, a puzzle game introduced in the 96 s, requires spatial awareness and an ability to work with
Lecture 3 - Regression
Lecture 3 - Regression Instructor: Prof Ganesh Ramakrishnan July 25, 2016 1 / 30 The Simplest ML Problem: Least Square Regression Curve Fitting: Motivation Error measurement Minimizing Error Method of
Time-average constraints in stochastic Model Predictive Control
Time-average constraints in stochastic Model Predictive Control James Fleming Mark Cannon ACC, May 2017 James Fleming, Mark Cannon Time-average constraints in stochastic MPC ACC, May 2017 1 / 24 Outline
Are there alternatives to Sigmoid Hidden Units? MLP Lecture 6 Hidden Units / Initialisation 1
Are there alternatives to Sigmoid Hidden Units? MLP Lecture 6 Hidden Units / Initialisation 1 Hidden Unit Transfer Functions Initialising Deep Networks Steve Renals Machine Learning Practical MLP Lecture
The Art of Neural Nets
The Art of Neural Nets Marco Tavora marcotav65@gmail.com Preamble The challenge of recognizing artists given their paintings has been, for a long time, far beyond the capability of algorithms. Recent advances
Adversarial Robustness for Aligned AI
Adversarial Robustness for Aligned AI Ian Goodfellow, Staff Research NIPS 2017 Workshop on Aligned Artificial Intelligence Many thanks to Catherine Olsson for feedback on drafts The Alignment Problem (This
Deep Learning for Launching and Mitigating Wireless Jamming Attacks
Deep Learning for Launching and Mitigating Wireless Jamming Attacks Tugba Erpek, Yalin E. Sagduyu, and Yi Shi arxiv:1807.02567v2 [cs.ni] 13 Dec 2018 Abstract An adversarial machine learning approach is
CAT Training CNNs for Image Classification with Noisy Labels
yclic Annealing Training (AT) NNs for Image lassification with Noisy Labels JiaWei Li, Tao Dai, QingTao Tang, YeLi Xing, Shu-Tao Xia Tsinghua University li-jw15@mailstsinghuaeducn October 8, 2018 AT Training
AUGMENTED CONVOLUTIONAL FEATURE MAPS FOR ROBUST CNN-BASED CAMERA MODEL IDENTIFICATION. Belhassen Bayar and Matthew C. Stamm
AUGMENTED CONVOLUTIONAL FEATURE MAPS FOR ROBUST CNN-BASED CAMERA MODEL IDENTIFICATION Belhassen Bayar and Matthew C. Stamm Department of Electrical and Computer Engineering, Drexel University, Philadelphia,
Chapter 2 Distributed Consensus Estimation of Wireless Sensor Networks
Chapter 2 Distributed Consensus Estimation of Wireless Sensor Networks Recently, consensus based distributed estimation has attracted considerable attention from various fields to estimate deterministic
Detection of Obscured Targets: Signal Processing
Detection of Obscured Targets: Signal Processing James McClellan and Waymond R. Scott, Jr. School of Electrical and Computer Engineering Georgia Institute of Technology Atlanta, GA 30332-0250 jim.mcclellan@ece.gatech.edu
Supplemental material of Robust Physical-World Attacks on Deep Learning Visual Classification"
Supplemental material of Robust Physical-World Attacks on Deep Learning Visual Classification" Kevin Eykholt 1, Ivan Evtimov *2, Earlence Fernandes 2, Bo Li 3, Amir Rahmati 4, Chaowei Xiao 1, Atul Prakash
Comparison of Google Image Search and ResNet Image Classification Using Image Similarity Metrics
University of Arkansas, Fayetteville ScholarWorks@UARK Computer Science and Computer Engineering Undergraduate Honors Theses Computer Science and Computer Engineering 5-2018 Comparison of Google Image
arxiv: v2 [cs.lg] 7 May 2017
![arxiv: v2 [cs.lg] 7 May 2017](/thumbs/74/71248397.jpg "arxiv: v2 [cs.lg] 7 May 2017")
STYLE TRANSFER GENERATIVE ADVERSARIAL NET- WORKS: LEARNING TO PLAY CHESS DIFFERENTLY Muthuraman Chidambaram & Yanjun Qi Department of Computer Science University of Virginia Charlottesville, VA 22903,
INFORMATION about image authenticity can be used in
1 Constrained Convolutional Neural Networs: A New Approach Towards General Purpose Image Manipulation Detection Belhassen Bayar, Student Member, IEEE, and Matthew C. Stamm, Member, IEEE Abstract Identifying
Playing CHIP-8 Games with Reinforcement Learning
Playing CHIP-8 Games with Reinforcement Learning Niven Achenjang, Patrick DeMichele, Sam Rogers Stanford University Abstract We begin with some background in the history of CHIP-8 games and the use of
Deep Learning for Autonomous Driving
Deep Learning for Autonomous Driving Shai Shalev-Shwartz Mobileye IMVC dimension, March, 2016 S. Shalev-Shwartz is also affiliated with The Hebrew University Shai Shalev-Shwartz (MobilEye) DL for Autonomous
Deep Learning Overview
Deep Learning Overview Eliu Huerta Gravity Group gravity.ncsa.illinois.edu National Center for Supercomputing Applications Department of Astronomy University of Illinois at Urbana-Champaign Data Visualization
Radio Deep Learning Efforts Showcase Presentation
Radio Deep Learning Efforts Showcase Presentation November 2016 hume@vt.edu www.hume.vt.edu Tim O Shea Senior Research Associate Program Overview Program Objective: Rethink fundamental approaches to how
Multimedia Forensics
Multimedia Forensics Using Mathematics and Machine Learning to Determine an Image's Source and Authenticity Matthew C. Stamm Multimedia & Information Security Lab (MISL) Department of Electrical and Computer
Classification of Road Images for Lane Detection
Classification of Road Images for Lane Detection Mingyu Kim minkyu89@stanford.edu Insun Jang insunj@stanford.edu Eunmo Yang eyang89@stanford.edu 1. Introduction In the research on autonomous car, it is
Review Sheet for Math 230, Midterm exam 2. Fall 2006
Review Sheet for Math 230, Midterm exam 2. Fall 2006 October 31, 2006 The second midterm exam will take place: Monday, November 13, from 8:15 to 9:30 pm. It will cover chapter 15 and sections 16.1 16.4,
Classification Accuracies of Malaria Infected Cells Using Deep Convolutional Neural Networks Based on Decompressed Images
Classification Accuracies of Malaria Infected Cells Using Deep Convolutional Neural Networks Based on Decompressed Images Yuhang Dong, Zhuocheng Jiang, Hongda Shen, W. David Pan Dept. of Electrical & Computer
Convolutional neural networks
Convolutional neural networks Themes Curriculum: Ch 9.1, 9.2 and http://cs231n.github.io/convolutionalnetworks/ The simple motivation and idea How it s done Receptive field Pooling Dilated convolutions
Compressive Through-focus Imaging
PIERS ONLINE, VOL. 6, NO. 8, 788 Compressive Through-focus Imaging Oren Mangoubi and Edwin A. Marengo Yale University, USA Northeastern University, USA Abstract Optical sensing and imaging applications
A Primer on Image Segmentation. Jonas Actor
A Primer on Image Segmentation It s all PDE s anyways Jonas Actor Rice University 21 February 2018 Jonas Actor Segmentation 21 February 2018 1 Table of Contents 1 Motivation 2 Simple Methods 3 Edge Methods
CS 4501: Introduction to Computer Vision. Filtering and Edge Detection
CS 451: Introduction to Computer Vision Filtering and Edge Detection Connelly Barnes Slides from Jason Lawrence, Fei Fei Li, Juan Carlos Niebles, Misha Kazhdan, Allison Klein, Tom Funkhouser, Adam Finkelstein,
tsushi Sasaki Fig. Flow diagram of panel structure recognition by specifying peripheral regions of each component in rectangles, and 3 types of detect
RECOGNITION OF NEL STRUCTURE IN COMIC IMGES USING FSTER R-CNN Hideaki Yanagisawa Hiroshi Watanabe Graduate School of Fundamental Science and Engineering, Waseda University BSTRCT For efficient e-comics
A TWO-PART PREDICTIVE CODER FOR MULTITASK SIGNAL COMPRESSION. Scott Deeann Chen and Pierre Moulin
A TWO-PART PREDICTIVE CODER FOR MULTITASK SIGNAL COMPRESSION Scott Deeann Chen and Pierre Moulin University of Illinois at Urbana-Champaign Department of Electrical and Computer Engineering 5 North Mathews
Kernels and Support Vector Machines
Kernels and Support Vector Machines Machine Learning CSE446 Sham Kakade University of Washington November 1, 2016 2016 Sham Kakade 1 Announcements: Project Milestones coming up HW2 You ve implemented GD,
Derek Allman a, Austin Reiter b, and Muyinatu Bell a,c
Exploring the effects of transducer models when training convolutional neural networks to eliminate reflection artifacts in experimental photoacoustic images Derek Allman a, Austin Reiter b, and Muyinatu
Machine Learning for Antenna Array Failure Analysis
Machine Learning for Antenna Array Failure Analysis Lydia de Lange Under Dr DJ Ludick and Dr TL Grobler Dept. Electrical and Electronic Engineering, Stellenbosch University MML 2019 Outline 15/03/2019
Automated Planetary Terrain Mapping of Mars Using Image Pattern Recognition
Automated Planetary Terrain Mapping of Mars Using Image Pattern Recognition Design Document Version 2.0 Team Strata: Sean Baquiro Matthew Enright Jorge Felix Tsosie Schneider 2 Table of Contents 1 Introduction.3
Learning to Predict Indoor Illumination from a Single Image. Chih-Hui Ho
Learning to Predict Indoor Illumination from a Single Image Chih-Hui Ho 1 Outline Introduction Method Overview LDR Panorama Light Source Detection Panorama Recentering Warp Learning From LDR Panoramas
Consistent Comic Colorization with Pixel-wise Background Classification
Consistent Comic Colorization with Pixel-wise Background Classification Sungmin Kang KAIST Jaegul Choo Korea University Jaehyuk Chang NAVER WEBTOON Corp. Abstract Comic colorization is a time-consuming
Convolutional Neural Networks
Convolutional Neural Networks Convolution, LeNet, AlexNet, VGGNet, GoogleNet, Resnet, DenseNet, CAM, Deconvolution Sept 17, 2018 Aaditya Prakash Convolution Convolution Demo Convolution Convolution in
Guido Cantelmo Prof. Francesco Viti. Practical methods for Dynamic Demand Estimation in congested Networks
Guido Cantelmo Prof. Francesco Viti MobiLab Transport Research Group Faculty of Sciences, Technology and Communication, Practical methods for Dynamic Demand Estimation in congested Networks University
Deep Learning for Infrastructure Assessment in Africa using Remote Sensing Data
Deep Learning for Infrastructure Assessment in Africa using Remote Sensing Data Pascaline Dupas Department of Economics, Stanford University Data for Development Initiative @ Stanford Center on Global
Lecture 23 Deep Learning: Segmentation
Lecture 23 Deep Learning: Segmentation COS 429: Computer Vision Thanks: most of these slides shamelessly adapted from Stanford CS231n: Convolutional Neural Networks for Visual Recognition Fei-Fei Li, Andrej
The Influence of Image Enhancement Filters on a Watermark Detection Rate Authors
acta graphica 194 udc 004.056.55:655.36 original scientific paper received: -09-011 accepted: 11-11-011 The Influence of Image Enhancement Filters on a Watermark Detection Rate Authors Ante Poljičak, Lidija
11/13/18. Introduction to RNNs for NLP. About Me. Overview SHANG GAO
Introduction to RNNs for NLP SHANG GAO About Me PhD student in the Data Science and Engineering program Took Deep Learning last year Work in the Biomedical Sciences, Engineering, and Computing group at
Practice problems from old exams for math 233
Practice problems from old exams for math 233 William H. Meeks III January 14, 2010 Disclaimer: Your instructor covers far more materials that we can possibly fit into a four/five questions exams. These
Stanford Center for AI Safety
Stanford Center for AI Safety Clark Barrett, David L. Dill, Mykel J. Kochenderfer, Dorsa Sadigh 1 Introduction Software-based systems play important roles in many areas of modern life, including manufacturing,
CLASSLESS ASSOCIATION USING NEURAL NETWORKS
Workshop track - ICLR 1 CLASSLESS ASSOCIATION USING NEURAL NETWORKS Federico Raue 1,, Sebastian Palacio, Andreas Dengel 1,, Marcus Liwicki 1 1 University of Kaiserslautern, Germany German Research Center
CS 4700: Foundations of Artificial Intelligence
CS 4700: Foundations of Artificial Intelligence selman@cs.cornell.edu Module: Adversarial Search R&N: Chapter 5 1 Outline Adversarial Search Optimal decisions Minimax α-β pruning Case study: Deep Blue
Game Theory for Safety and Security. Arunesh Sinha
Game Theory for Safety and Security Arunesh Sinha Motivation: Real World Security Issues 2 Central Problem Allocating limited security resources against an adaptive, intelligent adversary 3 Prior Work
Sensitivity study for the optimization of the viewing chord arrangement of the ITER poloidal polarimeter
P8-29 6th International Toki Conference, December 5-8, 26 Sensitivity study for the optimization of the viewing chord arrangement of the ITER poloidal polarimeter T. Yamaguchi, Y. Kawano and Y. Kusama
SYDE 112, LECTURE 34 & 35: Optimization on Restricted Domains and Lagrange Multipliers
SYDE 112, LECTURE 34 & 35: Optimization on Restricted Domains and Lagrange Multipliers 1 Restricted Domains If we are asked to determine the maximal and minimal values of an arbitrary multivariable function
Application of Multi Layer Perceptron (MLP) for Shower Size Prediction
Chapter 3 Application of Multi Layer Perceptron (MLP) for Shower Size Prediction 3.1 Basic considerations of the ANN Artificial Neural Network (ANN)s are non- parametric prediction tools that can be used
THE popularization of imaging components equipped in
IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY, VOL. 10, NO. 3, MARCH 2015 Revealing the Trace of High-Quality JPEG Compression Through Quantization Noise Analysis Bin Li, Member, IEEE, Tian-Tsong
Vehicle Detection using Images from Traffic Security Camera
Vehicle Detection using Images from Traffic Security Camera Lamia Iftekhar Final Report of Course Project CS174 May 30, 2012 1 1 The Task This project is an application of supervised learning algorithms.
Reliable Classification of Partially Occluded Coins
Reliable Classification of Partially Occluded Coins e-mail: L.J.P. van der Maaten P.J. Boon MICC, Universiteit Maastricht P.O. Box 616, 6200 MD Maastricht, The Netherlands telephone: (+31)43-3883901 fax:
Biologically Inspired Computation
Biologically Inspired Computation Deep Learning & Convolutional Neural Networks Joe Marino biologically inspired computation biological intelligence flexible capable of detecting/ executing/reasoning about
TD-Leaf(λ) Giraffe: Using Deep Reinforcement Learning to Play Chess. Stefan Lüttgen
TD-Leaf(λ) Giraffe: Using Deep Reinforcement Learning to Play Chess Stefan Lüttgen Motivation Learn to play chess Computer approach different than human one Humans search more selective: Kasparov (3-5
CC4.5: cost-sensitive decision tree pruning
Data Mining VI 239 CC4.5: cost-sensitive decision tree pruning J. Cai 1,J.Durkin 1 &Q.Cai 2 1 Department of Electrical and Computer Engineering, University of Akron, U.S.A. 2 Department of Electrical Engineering
Lecture 7: Scene Text Detection and Recognition. Dr. Cong Yao Megvii (Face++) Researcher
Lecture 7: Scene Text Detection and Recognition Dr. Cong Yao Megvii (Face++) Researcher yaocong@megvii.com Outline Background and Introduction Conventional Methods Deep Learning Methods Datasets and Competitions
ARGUING THE SAFETY OF MACHINE LEARNING FOR HIGHLY AUTOMATED DRIVING USING ASSURANCE CASES LYDIA GAUERHOF BOSCH CORPORATE RESEARCH
ARGUING THE SAFETY OF MACHINE LEARNING FOR HIGHLY AUTOMATED DRIVING USING ASSURANCE CASES 14.12.2017 LYDIA GAUERHOF BOSCH CORPORATE RESEARCH Arguing Safety of Machine Learning for Highly Automated Driving
To Post or Not To Post: Using CNNs to Classify Social Media Worthy Images
To Post or Not To Post: Using CNNs to Classify Social Media Worthy Images Lauren Blake Stanford University lblake@stanford.edu Abstract This project considers the feasibility for CNN models to classify
Strategic Network Formation with Structural Hole in R&D Projects: The Case Study on Japanese Cosmetic Industry
Journal of Robotics, Networking and Artificial Life, Vol. 3, No. 3 (December 2016), 188-192 Strategic Network Formation with Structural Hole in R&D Projects: The Case Study on Japanese Cosmetic Industry
Detection and Segmentation. Fei-Fei Li & Justin Johnson & Serena Yeung. Lecture 11 -
Lecture 11: Detection and Segmentation Lecture 11-1 May 10, 2017 Administrative Midterms being graded Please don t discuss midterms until next week - some students not yet taken A2 being graded Project
Class-count Reduction Techniques for Content Adaptive Filtering
Class-count Reduction Techniques for Content Adaptive Filtering Hao Hu Eindhoven University of Technology Eindhoven, the Netherlands Email: h.hu@tue.nl Gerard de Haan Philips Research Europe Eindhoven,
Land Cover Analysis to Determine Areas of Clear-cut and Forest Cover in Olney, Montana. Geob 373 Remote Sensing. Dr Andreas Varhola, Kathry De Rego
1 Land Cover Analysis to Determine Areas of Clear-cut and Forest Cover in Olney, Montana Geob 373 Remote Sensing Dr Andreas Varhola, Kathry De Rego Zhu an Lim (14292149) L2B 17 Apr 2016 2 Abstract Montana
Generating an appropriate sound for a video using WaveNet.
Australian National University College of Engineering and Computer Science Master of Computing Generating an appropriate sound for a video using WaveNet. COMP 8715 Individual Computing Project Taku Ueki
Lecture 11-1 CNN introduction. Sung Kim
Lecture 11-1 CNN introduction Sung Kim 'The only limit is your imagination' http://itchyi.squarespace.com/thelatest/2012/5/17/the-only-limit-is-your-imagination.html Lecture 7: Convolutional
The Wireless Network Jamming Problem Subject to Protocol Interference
The Wireless Network Jamming Problem Subject to Protocol Interference Author information blinded December 22, 2014 Abstract We study the following problem in wireless network security: Which jamming device
DIFFERENTIAL EQUATIONS. A principal model of physical phenomena.
DIFFERENTIAL EQUATIONS A principal model of physical phenomena. The ordinary differential equation: The initial value: y = f(x, y) y(x 0 )=Y 0 Find a solution Y (x) onsomeintervalx 0 x b. Together these
Background suppression with neural networks at the Belle II trigger
Background suppression with neural networks at the Belle II trigger Sebastian Skambraks Max-Planck-Institut für Physik March 28, 2017 Outline Introduction Motivation Trigger NeuroTrigger Background Neuro
Predicting when seam carved images become. unrecognizable. Sam Cunningham
Predicting when seam carved images become unrecognizable Sam Cunningham April 29, 2008 Acknowledgements I would like to thank my advisors, Shriram Krishnamurthi and Michael Tarr for all of their help along
JICE: Joint Data Compression and Encryption for Wireless Energy Auditing Networks
JICE: Joint Data Compression and Encryption for Wireless Energy Auditing Networks Sheng-Yuan Chiu 1,2, Hoang Hai Nguyen 1, Rui Tan 1, David K.Y. Yau 1,3,Deokwoo Jung 1 1 Advanced Digital Science Center,
Super resolution with Epitomes
Super resolution with Epitomes Aaron Brown University of Wisconsin Madison, WI Abstract Techniques exist for aligning and stitching photos of a scene and for interpolating image data to generate higher
6. Convolutional Neural Networks
6. Convolutional Neural Networks CS 519 Deep Learning, Winter 2016 Fuxin Li With materials from Zsolt Kira Quiz coming up Next Tuesday (1/26) 15 minutes Topics: Optimization Basic neural networks No Convolutional
RELEASING APERTURE FILTER CONSTRAINTS
RELEASING APERTURE FILTER CONSTRAINTS Jakub Chlapinski 1, Stephen Marshall 2 1 Department of Microelectronics and Computer Science, Technical University of Lodz, ul. Zeromskiego 116, 90-924 Lodz, Poland
Visualizing and Understanding. Fei-Fei Li & Justin Johnson & Serena Yeung. Lecture 12 -
Lecture 12: Visualizing and Understanding Lecture 12-1 May 16, 2017 Administrative Milestones due tonight on Canvas, 11:59pm Midterm grades released on Gradescope this week A3 due next Friday, 5/26 HyperQuest
Scrabble Board Automatic Detector for Third Party Applications
Scrabble Board Automatic Detector for Third Party Applications David Hirschberg Computer Science Department University of California, Irvine hirschbd@uci.edu Abstract Abstract Scrabble is a well-known
Vehicle Color Recognition using Convolutional Neural Network
Vehicle Color Recognition using Convolutional Neural Network Reza Fuad Rachmadi and I Ketut Eddy Purnama Multimedia and Network Engineering Department, Institut Teknologi Sepuluh Nopember, Keputih Sukolilo,
Swing Copters AI. Monisha White and Nolan Walsh Fall 2015, CS229, Stanford University
Swing Copters AI Monisha White and Nolan Walsh mewhite@stanford.edu njwalsh@stanford.edu Fall 2015, CS229, Stanford University 1. Introduction For our project we created an autonomous player for the game
Performance Analysis of a 1-bit Feedback Beamforming Algorithm
Performance Analysis of a 1-bit Feedback Beamforming Algorithm Sherman Ng Mark Johnson Electrical Engineering and Computer Sciences University of California at Berkeley Technical Report No. UCB/EECS-2009-161
Master s Thesis Defense
Master s Thesis Defense Comparison of Noncoherent Detectors for SOQPSK and GMSK in Phase Noise Channels Afzal Syed August 17, 2007 Committee Dr. Erik Perrins (Chair) Dr. Glenn Prescott Dr. Daniel Deavours
Deep Neural Network Architectures for Modulation Classification
Deep Neural Network Architectures for Modulation Classification Xiaoyu Liu, Diyu Yang, and Aly El Gamal School of Electrical and Computer Engineering Purdue University Email: {liu1962, yang1467, elgamala}@purdue.edu
TRANSFORMING PHOTOS TO COMICS USING CONVOLUTIONAL NEURAL NETWORKS. Tsinghua University, China Cardiff University, UK
TRANSFORMING PHOTOS TO COMICS USING CONVOUTIONA NEURA NETWORKS Yang Chen Yu-Kun ai Yong-Jin iu Tsinghua University, China Cardiff University, UK ABSTRACT In this paper, inspired by Gatys s recent work,
Color Constancy Using Standard Deviation of Color Channels
2010 International Conference on Pattern Recognition Color Constancy Using Standard Deviation of Color Channels Anustup Choudhury and Gérard Medioni Department of Computer Science University of Southern
Summary Overview of Topics in Econ 30200b: Decision theory: strong and weak domination by randomized strategies, domination theorem, expected utility
Summary Overview of Topics in Econ 30200b: Decision theory: strong and weak domination by randomized strategies, domination theorem, expected utility theorem (consistent decisions under uncertainty should