Thrun, Sebastian


Skin Cancer Detection and Tracking Using Data Synthesis and Deep Learning

AAAI Conferences

Dense object detection and temporal tracking are needed across applications domains ranging from people-tracking to analysis of satellite imagery over time. The detection and tracking of malignant skin cancers and benign moles poses a particularly challenging problem due to the general uniformity of large skin patches, the fact that skin lesions vary little in their appearance, and the relatively small amount of data available. Here we introduce a novel data synthesis technique that merges images of individual skin lesions with full-body images and heavily augments them to generate significant amounts of data. We build a convolutional neural network (CNN) based system, trained on this synthetic data, and demonstrate superior performance to traditional detection and tracking techniques. Additionally, we compare our system to humans trained with simple criteria. Our system is intended for potential clinical use to augment the capabilities of healthcare providers. While domain-specific, we believe the methods invoked in this work will be useful in applying CNNs across domains that suffer from limited data availability.


Learning to Track at 100 FPS with Deep Regression Networks

arXiv.org Artificial Intelligence

Machine learning techniques are often used in computer vision due to their ability to leverage large amounts of training data to improve performance. Unfortunately, most generic object trackers are still trained from scratch online and do not benefit from the large number of videos that are readily available for offline training. We propose a method for offline training of neural networks that can track novel objects at test-time at 100 fps. Our tracker is significantly faster than previous methods that use neural networks for tracking, which are typically very slow to run and not practical for real-time applications. Our tracker uses a simple feed-forward network with no online training required. The tracker learns a generic relationship between object motion and appearance and can be used to track novel objects that do not appear in the training set. We test our network on a standard tracking benchmark to demonstrate our tracker's state-of-the-art performance. Further, our performance improves as we add more videos to our offline training set. To the best of our knowledge, our tracker is the first neural-network tracker that learns to track generic objects at 100 fps.


A Bayesian Multiresolution Independence Test for Continuous Variables

arXiv.org Artificial Intelligence

In this paper we present a method ofcomputing the posterior probability ofconditional independence of two or morecontinuous variables from data,examined at several resolutions. Ourapproach is motivated by theobservation that the appearance ofcontinuous data varies widely atvarious resolutions, producing verydifferent independence estimatesbetween the variablesinvolved. Therefore, it is difficultto ascertain independence withoutexamining data at several carefullyselected resolutions. In our paper, weaccomplish this using the exactcomputation of the posteriorprobability of independence, calculatedanalytically given a resolution. Ateach examined resolution, we assume amultinomial distribution with Dirichletpriors for the discretized tableparameters, and compute the posteriorusing Bayesian integration. Acrossresolutions, we use a search procedureto approximate the Bayesian integral ofprobability over an exponential numberof possible histograms. Our methodgeneralizes to an arbitrary numbervariables in a straightforward manner.The test is suitable for Bayesiannetwork learning algorithms that useindependence tests to infer the networkstructure, in domains that contain anymix of continuous, ordinal andcategorical variables.


Particle Filters in Robotics (Invited Talk)

arXiv.org Artificial Intelligence

This presentation will introduce the audience to a new, emerging body of research on sequential Monte Carlo techniques in robotics. In recent years, particle filters have solved several hard perceptual robotic problems. Early successes were limited to low-dimensional problems, such as the problem of robot localization in environments with known maps. More recently, researchers have begun exploiting structural properties of robotic domains that have led to successful particle filter applications in spaces with as many as 100,000 dimensions. The presentation will discuss specific tricks necessary to make these techniques work in real - world domains,and also discuss open challenges for researchers IN the UAI community.


Learning Hierarchical Object Maps Of Non-Stationary Environments with mobile robots

arXiv.org Machine Learning

Building models, or maps, of robot environments is a highly active research area; however, most existing techniques construct unstructured maps and assume static environments. In this paper, we present an algorithm for learning object models of non-stationary objects found in office-type environments. Our algorithm exploits the fact that many objects found in office environments look alike (e.g., chairs, recycling bins). It does so through a two-level hierarchical representation, which links individual objects with generic shape templates of object classes. We derive an approximate EM algorithm for learning shape parameters at both levels of the hierarchy, using local occupancy grid maps for representing shape. Additionally, we develop a Bayesian model selection algorithm that enables the robot to estimate the total number of objects and object templates in the environment. Experimental results using a real robot equipped with a laser range finder indicate that our approach performs well at learning object-based maps of simple office environments. The approach outperforms a previously developed non-hierarchical algorithm that models objects but lacks class templates.


Decentralized Sensor Fusion With Distributed Particle Filters

arXiv.org Artificial Intelligence

This paper presents a scalable Bayesian technique for decentralized state estimation from multiple platforms in dynamic environments. As has long been recognized, centralized architectures impose severe scaling limitations for distributed systems due to the enormous communication overheads. We propose a strictly decentralized approach in which only nearby platforms exchange information. They do so through an interactive communication protocol aimed at maximizing information flow. Our approach is evaluated in the context of a distributed surveillance scenario that arises in a robotic system for playing the game of laser tag. Our results, both from simulation and using physical robots, illustrate an unprecedented scaling capability to large teams of vehicles.


Policy-contingent abstraction for robust robot control

arXiv.org Artificial Intelligence

This paper presents a scalable control algorithm that enables a deployed mobile robot system to make high-level decisions under full consideration of its probabilistic belief. Our approach is based on insights from the rich literature of hierarchical controllers and hierarchical MDPs. The resulting controller has been successfully deployed in a nursing facility near Pittsburgh, PA. To the best of our knowledge, this work is a unique instance of applying POMDPs to high-level robotic control problems.


Assisted Highway Lane Changing with RASCL

AAAI Conferences

Lane changing on highways is stressful. In this paper, we present RASCL, the Robotic Assistance System for Changing Lanes. RASCL combines state-of-the-art sensing and localization techniques with an accurate map describing road structure to detect and track other cars, determine whether or not a lane change to either side is safe, and communicate these safety statuses to the user using a variety of audio and visual interfaces. The user can interact with the system through specifying the size of their “comfort zone”, engaging the turn signal, or by simply driving across lane dividers. Additionally, RASCL provides speed change recommendations that are predicted to turn an unsafe lane change situation into a safe situation and enables communication with other vehicles by automatically controlling the turn signal when the driver attempts to change lanes without using the turn signal.


A Personal Account of the Development of Stanley, the Robot That Won the DARPA Grand Challenge

AI Magazine

This article is my personal account on the work at Stanford on Stanley, the winning robot in the DARPA Grand Challenge. Between July 2004 and October 2005, my then-postdoc Michael Montemerlo and I led a team of students, engineers, and professionals with the single vision of claiming one of the most prestigious trophies in the field of robotics: the DARPA Grand Challenge (DARPA 2004). The Grand Challenge, organized by the U.S. government, was unprecedented in the nation's history. Instead, this is my personal story of leading the Stanford Racing Team.


A Personal Account of the Development of Stanley, the Robot That Won the DARPA Grand Challenge

AI Magazine

This article is my personal account on the work at Stanford on Stanley, the winning robot in the DARPA Grand Challenge. Between July 2004 and October 2005, my then-postdoc Michael Montemerlo and I led a team of students, engineers, and professionals with the single vision of claiming one of the most prestigious trophies in the field of robotics: the DARPA Grand Challenge (DARPA 2004). The Grand Challenge, organized by the U.S. government, was unprecedented in the nation's history. It was the first time that the U.S. Congress had appropriated a cash price for advancing technological innovation. My team won this prize, competing with some 194 other teams. Stanley was the fastest of five robotic vehicles that, on October 8, 2005, successfully navigated a 131.6-mile-long course through California's Mojave Desert. This essay is not about the technology behind our success; for that I refer the interested reader to recent articles on the technical aspects of Stanley. Instead, this is my personal story of leading the Stanford Racing Team. It is the story of a team of people who built an autonomous robot in record time. It is also a success story for the field of artificial intelligence, as Stanley used some state of the art AI methods in areas such as probabilistic inference, machine learning, and computer vision. Of course, it is also the story of a step towards a technology that, one day, might fundamentally change our lives.