Models, code, and papers for "Benjamin Lee":

Defending Against Model Stealing Attacks Using Deceptive Perturbations

Sep 19, 2018
Taesung Lee, Benjamin Edwards, Ian Molloy, Dong Su

Machine learning models are vulnerable to simple model stealing attacks if the adversary can obtain output labels for chosen inputs. To protect against these attacks, it has been proposed to limit the information provided to the adversary by omitting probability scores, significantly impacting the utility of the provided service. In this work, we illustrate how a service provider can still provide useful, albeit misleading, class probability information, while significantly limiting the success of the attack. Our defense forces the adversary to discard the class probabilities, requiring significantly more queries before they can train a model with comparable performance. We evaluate several attack strategies, model architectures, and hyperparameters under varying adversarial models, and evaluate the efficacy of our defense against the strongest adversary. Finally, we quantify the amount of noise injected into the class probabilities to mesure the loss in utility, e.g., adding 1.26 nats per query on CIFAR-10 and 3.27 on MNIST. Our evaluation shows our defense can degrade the accuracy of the stolen model at least 20%, or require up to 64 times more queries while keeping the accuracy of the protected model almost intact.

* Under review for a peer review conference 

  Click for Model/Code and Paper
Gradient Descent Converges to Minimizers

Mar 04, 2016
Jason D. Lee, Max Simchowitz, Michael I. Jordan, Benjamin Recht

We show that gradient descent converges to a local minimizer, almost surely with random initialization. This is proved by applying the Stable Manifold Theorem from dynamical systems theory.

* Submitted to COLT 2016 

  Click for Model/Code and Paper
Efficient Exploration via State Marginal Matching

Jun 12, 2019
Lisa Lee, Benjamin Eysenbach, Emilio Parisotto, Eric Xing, Sergey Levine, Ruslan Salakhutdinov

To solve tasks with sparse rewards, reinforcement learning algorithms must be equipped with suitable exploration techniques. However, it is unclear what underlying objective is being optimized by existing exploration algorithms, or how they can be altered to incorporate prior knowledge about the task. Most importantly, it is difficult to use exploration experience from one task to acquire exploration strategies for another task. We address these shortcomings by learning a single exploration policy that can quickly solve a suite of downstream tasks in a multi-task setting, amortizing the cost of learning to explore. We recast exploration as a problem of State Marginal Matching (SMM): we learn a mixture of policies for which the state marginal distribution matches a given target state distribution, which can incorporate prior knowledge about the task. Without any prior knowledge, the SMM objective reduces to maximizing the marginal state entropy. We optimize the objective by reducing it to a two-player, zero-sum game, where we iteratively fit a state density model and then update the policy to visit states with low density under this model. While many previous algorithms for exploration employ a similar procedure, they omit a crucial historical averaging step, without which the iterative procedure does not converge to a Nash equilibria. To parallelize exploration, we extend our algorithm to use mixtures of policies, wherein we discover connections between SMM and previously-proposed skill learning methods based on mutual information. On complex navigation and manipulation tasks, we demonstrate that our algorithm explores faster and adapts more quickly to new tasks.

* Videos and code: 

  Click for Model/Code and Paper
First-order Methods Almost Always Avoid Saddle Points

Oct 20, 2017
Jason D. Lee, Ioannis Panageas, Georgios Piliouras, Max Simchowitz, Michael I. Jordan, Benjamin Recht

We establish that first-order methods avoid saddle points for almost all initializations. Our results apply to a wide variety of first-order methods, including gradient descent, block coordinate descent, mirror descent and variants thereof. The connecting thread is that such algorithms can be studied from a dynamical systems perspective in which appropriate instantiations of the Stable Manifold Theorem allow for a global stability analysis. Thus, neither access to second-order derivative information nor randomness beyond initialization is necessary to provably avoid saddle points.

  Click for Model/Code and Paper
Detecting Backdoor Attacks on Deep Neural Networks by Activation Clustering

Nov 09, 2018
Bryant Chen, Wilka Carvalho, Nathalie Baracaldo, Heiko Ludwig, Benjamin Edwards, Taesung Lee, Ian Molloy, Biplav Srivastava

While machine learning (ML) models are being increasingly trusted to make decisions in different and varying areas, the safety of systems using such models has become an increasing concern. In particular, ML models are often trained on data from potentially untrustworthy sources, providing adversaries with the opportunity to manipulate them by inserting carefully crafted samples into the training set. Recent work has shown that this type of attack, called a poisoning attack, allows adversaries to insert backdoors or trojans into the model, enabling malicious behavior with simple external backdoor triggers at inference time and only a blackbox perspective of the model itself. Detecting this type of attack is challenging because the unexpected behavior occurs only when a backdoor trigger, which is known only to the adversary, is present. Model users, either direct users of training data or users of pre-trained model from a catalog, may not guarantee the safe operation of their ML-based system. In this paper, we propose a novel approach to backdoor detection and removal for neural networks. Through extensive experimental results, we demonstrate its effectiveness for neural networks classifying text and images. To the best of our knowledge, this is the first methodology capable of detecting poisonous data crafted to insert backdoors and repairing the model that does not require a verified and trusted dataset.

  Click for Model/Code and Paper
Query Strategies for Evading Convex-Inducing Classifiers

Jul 03, 2010
Blaine Nelson, Benjamin I. P. Rubinstein, Ling Huang, Anthony D. Joseph, Steven J. Lee, Satish Rao, J. D. Tygar

Classifiers are often used to detect miscreant activities. We study how an adversary can systematically query a classifier to elicit information that allows the adversary to evade detection while incurring a near-minimal cost of modifying their intended malfeasance. We generalize the theory of Lowd and Meek (2005) to the family of convex-inducing classifiers that partition input space into two sets one of which is convex. We present query algorithms for this family that construct undetected instances of approximately minimal cost using only polynomially-many queries in the dimension of the space and in the level of approximation. Our results demonstrate that near-optimal evasion can be accomplished without reverse-engineering the classifier's decision boundary. We also consider general lp costs and show that near-optimal evasion on the family of convex-inducing classifiers is generally efficient for both positive and negative convexity for all levels of approximation if p=1.

  Click for Model/Code and Paper
An Approach for Self-Training Audio Event Detectors Using Web Data

Jun 27, 2017
Benjamin Elizalde, Ankit Shah, Siddharth Dalmia, Min Hun Lee, Rohan Badlani, Anurag Kumar, Bhiksha Raj, Ian Lane

Audio Event Detection (AED) aims to recognize sounds within audio and video recordings. AED employs machine learning algorithms commonly trained and tested on annotated datasets. However, available datasets are limited in number of samples and hence it is difficult to model acoustic diversity. Therefore, we propose combining labeled audio from a dataset and unlabeled audio from the web to improve the sound models. The audio event detectors are trained on the labeled audio and ran on the unlabeled audio downloaded from YouTube. Whenever the detectors recognized any of the known sounds with high confidence, the unlabeled audio was use to re-train the detectors. The performance of the re-trained detectors is compared to the one from the original detectors using the annotated test set. Results showed an improvement of the AED, and uncovered challenges of using web audio from videos.

* 5 pages 

  Click for Model/Code and Paper
Near-Optimal Evasion of Convex-Inducing Classifiers

Mar 14, 2010
Blaine Nelson, Benjamin I. P. Rubinstein, Ling Huang, Anthony D. Joseph, Shing-hon Lau, Steven J. Lee, Satish Rao, Anthony Tran, J. D. Tygar

Classifiers are often used to detect miscreant activities. We study how an adversary can efficiently query a classifier to elicit information that allows the adversary to evade detection at near-minimal cost. We generalize results of Lowd and Meek (2005) to convex-inducing classifiers. We present algorithms that construct undetected instances of near-minimal cost using only polynomially many queries in the dimension of the space and without reverse engineering the decision boundary.

* 8 pages; to appear at AISTATS'2010 

  Click for Model/Code and Paper
Large-Scale Video Classification with Feature Space Augmentation coupled with Learned Label Relations and Ensembling

Sep 21, 2018
Choongyeun Cho, Benjamin Antin, Sanchit Arora, Shwan Ashrafi, Peilin Duan, Dang The Huynh, Lee James, Hang Tuan Nguyen, Mojtaba Solgi, Cuong Van Than

This paper presents the Axon AI's solution to the 2nd YouTube-8M Video Understanding Challenge, achieving the final global average precision (GAP) of 88.733% on the private test set (ranked 3rd among 394 teams, not considering the model size constraint), and 87.287% using a model that meets size requirement. Two sets of 7 individual models belonging to 3 different families were trained separately. Then, the inference results on a training data were aggregated from these multiple models and fed to train a compact model that meets the model size requirement. In order to further improve performance we explored and employed data over/sub-sampling in feature space, an additional regularization term during training exploiting label relationship, and learned weights for ensembling different individual models.

  Click for Model/Code and Paper
Computing CNN Loss and Gradients for Pose Estimation with Riemannian Geometry

Jul 17, 2018
Benjamin Hou, Nina Miolane, Bishesh Khanal, Matthew C. H. Lee, Amir Alansary, Steven McDonagh, Jo V. Hajnal, Daniel Rueckert, Ben Glocker, Bernhard Kainz

Pose estimation, i.e. predicting a 3D rigid transformation with respect to a fixed co-ordinate frame in, SE(3), is an omnipresent problem in medical image analysis with applications such as: image rigid registration, anatomical standard plane detection, tracking and device/camera pose estimation. Deep learning methods often parameterise a pose with a representation that separates rotation and translation. As commonly available frameworks do not provide means to calculate loss on a manifold, regression is usually performed using the L2-norm independently on the rotation's and the translation's parameterisations, which is a metric for linear spaces that does not take into account the Lie group structure of SE(3). In this paper, we propose a general Riemannian formulation of the pose estimation problem. We propose to train the CNN directly on SE(3) equipped with a left-invariant Riemannian metric, coupling the prediction of the translation and rotation defining the pose. At each training step, the ground truth and predicted pose are elements of the manifold, where the loss is calculated as the Riemannian geodesic distance. We then compute the optimisation direction by back-propagating the gradient with respect to the predicted pose on the tangent space of the manifold SE(3) and update the network weights. We thoroughly evaluate the effectiveness of our loss function by comparing its performance with popular and most commonly used existing methods, on tasks such as image-based localisation and intensity-based 2D/3D registration. We also show that hyper-parameters, used in our loss function to weight the contribution between rotations and translations, can be intrinsically calculated from the dataset to achieve greater performance margins.

  Click for Model/Code and Paper
A Deep Dive into Understanding Tumor Foci Classification using Multiparametric MRI Based on Convolutional Neural Network

Apr 04, 2019
Weiwei Zong, Joon Lee, Chang Liu, Eric Carver, Aharon Feldman, Branislava Janic, Mohamed Elshaikh, Milan Pantelic, David Hearshen, Indrin Chetty, Benjamin Movsas, Ning Wen

Data scarcity has refrained deep learning models from making greater progress in prostate images analysis using multiparametric MRI. In this paper, an efficient convolutional neural network (CNN) was developed to classify lesion malignancy for prostate cancer patients, based on which model interpretation was systematically analyzed to bridge the gap between natural images and MR images, which is the first one of its kind in the literature. The problem of small sample size was addressed and successfully tackled by feeding the intermediate features into a traditional classification algorithm known as weighted extreme learning machine, with imbalanced distribution among output categories taken into consideration. Model trained on public data set was able to generalize to data from an independent institution to make accurate prediction. The generated saliency map was found to overlay well with the lesion and could benefit clinicians for diagnosing purpose.

  Click for Model/Code and Paper
Gmail Smart Compose: Real-Time Assisted Writing

May 17, 2019
Mia Xu Chen, Benjamin N Lee, Gagan Bansal, Yuan Cao, Shuyuan Zhang, Justin Lu, Jackie Tsay, Yinan Wang, Andrew M. Dai, Zhifeng Chen, Timothy Sohn, Yonghui Wu

In this paper, we present Smart Compose, a novel system for generating interactive, real-time suggestions in Gmail that assists users in writing mails by reducing repetitive typing. In the design and deployment of such a large-scale and complicated system, we faced several challenges including model selection, performance evaluation, serving and other practical issues. At the core of Smart Compose is a large-scale neural language model. We leveraged state-of-the-art machine learning techniques for language model training which enabled high-quality suggestion prediction, and constructed novel serving infrastructure for high-throughput and real-time inference. Experimental results show the effectiveness of our proposed system design and deployment approach. This system is currently being served in Gmail.

  Click for Model/Code and Paper
Project AutoVision: Localization and 3D Scene Perception for an Autonomous Vehicle with a Multi-Camera System

Mar 05, 2019
Lionel Heng, Benjamin Choi, Zhaopeng Cui, Marcel Geppert, Sixing Hu, Benson Kuan, Peidong Liu, Rang Nguyen, Ye Chuan Yeo, Andreas Geiger, Gim Hee Lee, Marc Pollefeys, Torsten Sattler

Project AutoVision aims to develop localization and 3D scene perception capabilities for a self-driving vehicle. Such capabilities will enable autonomous navigation in urban and rural environments, in day and night, and with cameras as the only exteroceptive sensors. The sensor suite employs many cameras for both 360-degree coverage and accurate multi-view stereo; the use of low-cost cameras keeps the cost of this sensor suite to a minimum. In addition, the project seeks to extend the operating envelope to include GNSS-less conditions which are typical for environments with tall buildings, foliage, and tunnels. Emphasis is placed on leveraging multi-view geometry and deep learning to enable the vehicle to localize and perceive in 3D space. This paper presents an overview of the project, and describes the sensor suite and current progress in the areas of calibration, localization, and perception.

* 2019 IEEE International Conference on Robotics and Automation (ICRA) 

  Click for Model/Code and Paper
Lingvo: a Modular and Scalable Framework for Sequence-to-Sequence Modeling

Feb 21, 2019
Jonathan Shen, Patrick Nguyen, Yonghui Wu, Zhifeng Chen, Mia X. Chen, Ye Jia, Anjuli Kannan, Tara Sainath, Yuan Cao, Chung-Cheng Chiu, Yanzhang He, Jan Chorowski, Smit Hinsu, Stella Laurenzo, James Qin, Orhan Firat, Wolfgang Macherey, Suyog Gupta, Ankur Bapna, Shuyuan Zhang, Ruoming Pang, Ron J. Weiss, Rohit Prabhavalkar, Qiao Liang, Benoit Jacob, Bowen Liang, HyoukJoong Lee, Ciprian Chelba, Sébastien Jean, Bo Li, Melvin Johnson, Rohan Anil, Rajat Tibrewal, Xiaobing Liu, Akiko Eriguchi, Navdeep Jaitly, Naveen Ari, Colin Cherry, Parisa Haghani, Otavio Good, Youlong Cheng, Raziel Alvarez, Isaac Caswell, Wei-Ning Hsu, Zongheng Yang, Kuan-Chieh Wang, Ekaterina Gonina, Katrin Tomanek, Ben Vanik, Zelin Wu, Llion Jones, Mike Schuster, Yanping Huang, Dehao Chen, Kazuki Irie, George Foster, John Richardson, Klaus Macherey, Antoine Bruguier, Heiga Zen, Colin Raffel, Shankar Kumar, Kanishka Rao, David Rybach, Matthew Murray, Vijayaditya Peddinti, Maxim Krikun, Michiel A. U. Bacchiani, Thomas B. Jablin, Rob Suderman, Ian Williams, Benjamin Lee, Deepti Bhatia, Justin Carlson, Semih Yavuz, Yu Zhang, Ian McGraw, Max Galkin, Qi Ge, Golan Pundak, Chad Whipkey, Todd Wang, Uri Alon, Dmitry Lepikhin, Ye Tian, Sara Sabour, William Chan, Shubham Toshniwal, Baohua Liao, Michael Nirschl, Pat Rondon

Lingvo is a Tensorflow framework offering a complete solution for collaborative deep learning research, with a particular focus towards sequence-to-sequence models. Lingvo models are composed of modular building blocks that are flexible and easily extensible, and experiment configurations are centralized and highly customizable. Distributed training and quantized inference are supported directly within the framework, and it contains existing implementations of a large number of utilities, helper functions, and the newest research ideas. Lingvo has been used in collaboration by dozens of researchers in more than 20 papers over the last two years. This document outlines the underlying design of Lingvo and serves as an introduction to the various pieces of the framework, while also offering examples of advanced features that showcase the capabilities of the framework.

  Click for Model/Code and Paper
Sequential Learning of Principal Curves: Summarizing Data Streams on the Fly

May 18, 2018
Benjamin Guedj, Le Li

When confronted with massive data streams, summarizing data with dimension reduction methods such as PCA raises theoretical and algorithmic pitfalls. Principal curves act as a nonlinear generalization of PCA and the present paper proposes a novel algorithm to automatically and sequentially learn principal curves from data streams. We show that our procedure is supported by regret bounds with optimal sublinear remainder terms. A greedy local search implementation that incorporates both sleeping experts and multi-armed bandit ingredients is presented, along with its regret bound and performance on a toy example and seismic data.

  Click for Model/Code and Paper
A Resource Allocation based Approach for Corporate Mobility as a Service

Oct 11, 2018
Sebastian Knopp, Benjamin Biesinger, Matthias Prandtstetter

Corporate mobility is often based on fixed assignments of vehicles to employees. Relaxing these fixed assignments while including alternatives such as public transportation, bike sharing, and taxis for the employees' business and private trips could increase fleet utilization, foster the use of battery electric vehicles, and lower the costs for the companies' transportation needs. A system in which all employees specify their mobility demands gives rise to optimization problems concerning the assignment of company cars or alternative modes of transport to satisfy the needs of the users. In this work we introduce the NP-hard mobility offer allocation problem which has similarities to interval scheduling problems. We propose an integer linear programming model and heuristic solution approaches based on large neighborhood search. The efficiency of these methods is based on the usage of suitable conflict graphs. In a computational study, the approaches are evaluated and it is demonstrated that, depending on instances and run-time requirements, either solving the model exactly using a general purpose integer linear programming solver, fast greedy heuristics, or the adaptive large neighborhood search outperforms the others.

  Click for Model/Code and Paper
Transfer Learning for Prosthetics Using Imitation Learning

Jan 15, 2019
Montaser Mohammedalamen, Waleed D. Khamies, Benjamin Rosman

In this paper, We Apply Reinforcement learning (RL) techniques to train a realistic biomechanical model to work with different people and on different walking environments. We benchmarking 3 RL algorithms: Deep Deterministic Policy Gradient (DDPG), Trust Region Policy Optimization (TRPO) and Proximal Policy Optimization (PPO) in OpenSim environment, Also we apply imitation learning to a prosthetics domain to reduce the training time needed to design customized prosthetics. We use DDPG algorithm to train an original expert agent. We then propose a modification to the Dataset Aggregation (DAgger) algorithm to reuse the expert knowledge and train a new target agent to replicate that behaviour in fewer than 5 iterations, compared to the 100 iterations taken by the expert agent which means reducing training time by 95%. Our modifications to the DAgger algorithm improve the balance between exploiting the expert policy and exploring the environment. We show empirically that these improve convergence time of the target agent, particularly when there is some degree of variation between expert and naive agent.

* Black in AI Workshop, NeurIPS 2018 
* Workshop paper, Black in AI, NeurIPS 2018 

  Click for Model/Code and Paper
RADMPC: A Fast Decentralized Approach for Chance-Constrained Multi-Vehicle Path-Planning

Nov 25, 2018
Aaron Huang, Benjamin J. Ayton, Brian C. Williams

Robust multi-vehicle path-planning is important for ensuring the safety of multi-vehicle systems in applications like transportation, search and rescue, and robotic exploration. Chance-constrained methods like Iterative Risk Allocation (IRA)\cite{IRA} have been developed for situations where environmental disturbances are unbounded. However, chance-constrained methods for the multi-vehicle case generally use centralized strategies where the vehicle set is planned with couplings between all vehicle pairs. This approach is intractable as fleet size increases because computation time is exponential with respect to the number of vehicles being planned over due to a polynomial increase in coupling constraints between vehicle pairs. We present a faster approach for chance-constrained multi-vehicle path-planning that relies upon a decentralized path-planning method called Risk-Aware Decentralized Model Predictive Control (RADMPC) to rapidly approximate a centralized IRA approach. The RADMPC approximation is evaluated for vehicle interactions to determine the vehicle sets that should be planned in a coupled manner. Applying IRA to the smaller vehicle sets determined from the RADMPC approximation rapidly plans safe paths for the entire fleet. A Monte Carlo simulation analysis demonstrates the correctness of our approach and a significant improvement in computation time compared to a centralized IRA approach.

  Click for Model/Code and Paper
Fog Robotics: A Summary, Challenges and Future Scope

Aug 14, 2019
Siva Leela Krishna Chand Gudi, Benjamin Johnston, Mary-Anne Williams

Human-robot interaction plays a crucial role to make robots closer to humans. Usually, robots are limited by their own capabilities. Therefore, they utilise Cloud Robotics to enhance their dexterity. Its ability includes the sharing of information such as maps, images and the processing power. This whole process involves distributing data which intend to rise enormously. New issues can arise such as bandwidth, network congestion at backhaul and fronthaul systems resulting in high latency. Thus, it can make an impact on seamless connectivity between the robots, users and the cloud. Also, a robot may not accomplish its goal successfully within a stipulated time. As a consequence, Cloud Robotics cannot be in a position to handle the traffic imposed by robots. On the contrary, impending Fog Robotics can act as a solution by solving major problems of Cloud Robotics. Therefore to check its feasibility, we discuss the need and architectures of Fog Robotics in this paper. To evaluate the architectures, we used a realistic scenario of Fog Robotics by comparing them with Cloud Robotics. Next, latency is chosen as the primary factor for validating the effectiveness of the system. Besides, we utilised real-time latency using Pepper robot, Fog robot server and the Cloud server. Experimental results show that Fog Robotics reduces latency significantly compared to Cloud Robotics. Moreover, advantages, challenges and future scope of the Fog Robotics system is further discussed.

* 10 pages, 6 figures 

  Click for Model/Code and Paper
"Good Robot!": Efficient Reinforcement Learning for Multi-Step Visual Tasks via Reward Shaping

Sep 25, 2019
Andrew Hundt, Benjamin Killeen, Heeyeon Kwon, Chris Paxton, Gregory D. Hager

In order to learn effectively, robots must be able to extract the intangible context by which task progress and mistakes are defined. In the domain of reinforcement learning, much of this information is provided by the reward function. Hence, reward shaping is a necessary part of how we can achieve state-of-the-art results on complex, multi-step tasks. However, comparatively little work has examined how reward shaping should be done so that it captures task context, particularly in scenarios where the task is long-horizon and failure is highly consequential. Our Schedule for Positive Task (SPOT) reward trains our Efficient Visual Task (EVT) model to solve problems that require an understanding of both task context and workspace constraints of multi-step block arrangement tasks. In simulation EVT can completely clear adversarial arrangements of objects by pushing and grasping in 99% of cases vs an 82% baseline in prior work. For random arrangements EVT clears 100% of test cases at 86% action efficiency vs 61% efficiency in prior work. EVT + SPOT is also able to demonstrate context understanding and complete stacks in 74% of trials compared to a baseline of 5% with EVT alone. To our knowledge, this is the first instance of a Reinforcement Learning based algorithm successfully completing such a challenge. Code is available at .

* 7 pages, 6 figures, code is available at 

  Click for Model/Code and Paper