Learning and Inferring Transportation Routines Lin Liao, Don Patterson, Dieter Fox, Henry Kautz Department of Computer Science and Engineering University.

Slides:



Advertisements
Similar presentations
MTA ETA. Product Description A real-time simulation system that estimates the expected time that it will take a certain bus to arrive at an end- users.
Advertisements

Opportunity Knocks: A Community Navigation Aid Henry Kautz Don Patterson Dieter Fox Lin Liao University of Washington Computer Science & Engineering.
An Interactive-Voting Based Map Matching Algorithm
Recognizing Human Activity from Sensor Data Henry Kautz University of Washington Computer Science & Engineering graduate students: Don Patterson, Lin Liao.
State Estimation and Kalman Filtering CS B659 Spring 2013 Kris Hauser.
Online Filtering, Smoothing & Probabilistic Modeling of Streaming Data In short, Applying probabilistic models to Streams Bhargav Kanagal & Amol Deshpande.
1 Assisted Cognition Henry Kautz Don Patterson, Nan LI Oren Etzioni, Dieter Fox University of Washington Department of Computer Science & Engineering.
Assisted Cognition Henry Kautz University of Rochester Computer Science.
Christian Mandel Bernd Krieg-Brückner Bernd Gersdorf Christoph Budelmann Marcus-Sebastian Schröder Navigation Aid for Mobility Assistants Joint CEWIT-TZI-acatech.
Chapter 15 Probabilistic Reasoning over Time. Chapter 15, Sections 1-5 Outline Time and uncertainty Inference: ltering, prediction, smoothing Hidden Markov.
1 Assisted Cognition Henry Kautz, Oren Etzioni, & Dieter Fox University of Washington Department of Computer Science & Engineering.
Individual Localization and Tracking in Multi-Robot Settings with Dynamic Landmarks Anousha Mesbah Prashant Doshi Prashant Doshi University of Georgia.
Constructing Popular Routes from Uncertain Trajectories Ling-Yin Wei 1, Yu Zheng 2, Wen-Chih Peng 1 1 National Chiao Tung University, Taiwan 2 Microsoft.
Transportation mode detection using mobile phones and GIS information Leon Stenneth, Ouri Wolfson, Philip Yu, Bo Xu 1University of Illinois, Chicago.
Dieter Fox University of Washington Department of Computer Science & Engineering.
TOWARD DYNAMIC GRASP ACQUISITION: THE G-SLAM PROBLEM Li (Emma) Zhang and Jeff Trinkle Department of Computer Science, Rensselaer Polytechnic Institute.
CS 547: Sensing and Planning in Robotics Gaurav S. Sukhatme Computer Science Robotic Embedded Systems Laboratory University of Southern California
Search and Rescue Optimal Planning System
Particle Filters for Mobile Robot Localization 11/24/2006 Aliakbar Gorji Roborics Instructor: Dr. Shiri Amirkabir University of Technology.
MTA ETA. Product Description A real-time simulation system that estimates the expected time that it will take a certain bus to arrive at an end- user’s.
Probabilistic Databases Amol Deshpande, University of Maryland.
Learning Programs Danielle and Joseph Bennett (and Lorelei) 4 December 2007.
Non-invasive Techniques for Human Fatigue Monitoring Qiang Ji Dept. of Electrical, Computer, and Systems Engineering Rensselaer Polytechnic Institute
Learning Transportation Mode from Raw GPS Data for Geographic Applications on the Web Yu Zheng, Like Liu, Xing Xie Microsoft Research.
[Context to Make You More Aware] Presentation [Adrienne Andrew, Yaw Anokwa, Karl Koscher, Jonathan Lester, Gaetano Borriello Department of Computer Science.
Understanding User Mobility Based on GPS Data Yu Zheng Microsoft Research Asia.
Exploration of Ground Truth from Raw GPS Data National University of Defense Technology & Hong Kong University of Science and Technology Exploration of.
HCI / CprE / ComS 575: Computational Perception
Personalized Medicine Research at the University of Rochester Henry Kautz Department of Computer Science.
Bayesian Filtering for Robot Localization
Sensys 2009 Speaker:Lawrence.  Introduction  Overview & Challenges  Algorithm  Travel Time Estimation  Evaluation  Conclusion.
GPS Trajectories Analysis in MOPSI Project Minjie Chen SIPU group Univ. of Eastern Finland.
1 GIS Transit Applications in Fairfax County, VA Brendan Ford Department of Information Technology (GIS) Wendy Jia Department of Transportation June, 1999.
6 am 11 am 5 pm Fig. 5: Population density estimates using the aggregated Markov chains. Colour scale represents people per km. Population Activity Estimation.
Presented by: Chaitanya K. Sambhara Paper by: Maarten Ditzel, Caspar Lageweg, Johan Janssen, Arne Theil TNO Defence, Security and Safety, The Hague, The.
Extracting Places and Activities from GPS Traces Using Hierarchical Conditional Random Fields Yong-Joong Kim Dept. of Computer Science Yonsei.
Markov Localization & Bayes Filtering
Robust Activity Recognition Henry Kautz University of Washington Computer Science & Engineering graduate students: Don Patterson, Lin Liao, Krzysztof Gajos,
Introduction to MCMC and BUGS. Computational problems More parameters -> even more parameter combinations Exact computation and grid approximation become.
From Bayesian Filtering to Particle Filters Dieter Fox University of Washington Joint work with W. Burgard, F. Dellaert, C. Kwok, S. Thrun.
Making Sense of Sensors Henry Kautz Department of Computer Science & Engineering University of Washington, Seattle, WA Funding for this research is provided.
GPS Calibrated Ad-hoc Localization for Geosocial Networking Dexter H. Hu Cho-Li Wang Yinfeng Wang
Mapping and Localization with RFID Technology Matthai Philipose, Kenneth P Fishkin, Dieter Fox, Dirk Hahnel, Wolfram Burgard Presenter: Aniket Shah.
The Science of Prediction Location Intelligence Conference April 4, 2006 How Next Generation Traffic Services Will Impact Business Dr. Oliver Downs, Chief.
Inferring High-Level Behavior from Low-Level Sensors Don Peterson, Lin Liao, Dieter Fox, Henry Kautz Published in UBICOMP 2003 ICS 280.
Multi-hop-based Monte Carlo Localization for Mobile Sensor Networks
Computer Science, Software Engineering & Robotics Workshop, FGCU, April 27-28, 2012 Fault Prediction with Particle Filters by David Hatfield mentors: Dr.
Learning and Inferring Transportation Routines By: Lin Liao, Dieter Fox and Henry Kautz Best Paper award AAAI’04.
Dynamic Bayesian Networks and Particle Filtering COMPSCI 276 (chapter 15, Russel and Norvig) 2007.
College of Engineering Anchor Nodes Placement for Effective Passive Localization Karthikeyan Pasupathy Major Advisor: Dr. Robert Akl Department of Computer.
Learning to Navigate Through Crowded Environments Peter Henry 1, Christian Vollmer 2, Brian Ferris 1, Dieter Fox 1 Tuesday, May 4, University of.
Tracking Multiple Cells By Correspondence Resolution In A Sequential Bayesian Framework Nilanjan Ray Gang Dong Scott T. Acton C.L. Brown Department of.
Inferring High-Level Behavior from Low-Level Sensors Donald J. Patterson, Lin Liao, Dieter Fox, and Henry Kautz.
Intelligent DataBase System Lab, NCKU, Taiwan Josh Jia-Ching Ying 1, Wang-Chien Lee 2, Tz-Chiao Weng 1 and Vincent S. Tseng 1 1 Department of Computer.
Student Name USN NO Guide Name H.O.D Name Name Of The College & Dept.
Expectation-Maximization (EM) Algorithm & Monte Carlo Sampling for Inference and Approximation.
HCI/ComS 575X: Computational Perception Instructor: Alexander Stoytchev
Project ACCESS Henry Kautz, Dieter Fox, Gaetano Boriello (UW CSE) Don Patterson (UW / UC Irvine) Kurt Johnson, Pat Brown, Mark Harniss (UW Rehabilitation.
Real-time Bus Information on Mobile Devices Stuart D. Maclean, Daniel J. Dailey ITS Research Program University of Washington, Seattle, WA.
Rao-Blackwellised Particle Filtering for Dynamic Bayesian Networks Arnaud Doucet, Nando de Freitas, Kevin Murphy and Stuart Russell CS497EA presentation.
1 Relational Factor Graphs Lin Liao Joint work with Dieter Fox.
Mobile Robotics. Fundamental Idea: Robot Pose 2D world (floor plan) 3 DOF Very simple model—the difficulty is in autonomy.
Trajectory Data Mining Dr. Yu Zheng Lead Researcher, Microsoft Research Chair Professor at Shanghai Jiao Tong University Editor-in-Chief of ACM Trans.
Assisted Cognition Systems Henry Kautz Department of Computer Science.
Mohsen Riahi Manesh and Dr. Naima Kaabouch
Understanding Human Behavior from Sensor Data
Probabilistic Data Management
People Forecasting Where people are going?
Markov Chain Monte Carlo
Finding Periodic Discrete Events in Noisy Streams
Presentation transcript:

Learning and Inferring Transportation Routines Lin Liao, Don Patterson, Dieter Fox, Henry Kautz Department of Computer Science and Engineering University of Washington

Outline Motivations and Overview  Global Positioning System (GPS)  Geographic Information System (GIS)  Probabilistic Reasoning Engine Probabilistic Reasoning Engine  Graphical model  Inference  Learning  Error detection Conclusion

Global Positioning System

GPS Receivers Garmin's iQue 3600 GPS-enabled PDA Samsung N300 GPS- enabled Cell Phone GeoStats wearable GPS logger

Geographic Information System Bus routes and bus stops Data source: Metro GIS Street map Data source: Census 2000 Tiger/line data

Motivations GPS, GIS High Level Activities (goals, paths, transportation modes, errors, etc)

Architecture Learning Engine Inference Engine GIS Database Probabilistic Model Goals Paths Modes Errors

Just-in-time Travel Information customized for each individual An ordinary real time traffic condition service Display traffic condition only on the predicted most likely trajectory

Activity Compass An intelligent PDA which helps guide a cognitively impaired person safely through the community. The system notes when the user departs from a familiar routine and provides proactive alerts or calls for assistance.

Probabilistic Reasoning Graphical model: Dynamic Bayesian Network Inference engine: Rao-Blackwellised particle filters Learning engine: Expectation-Maximization (EM) algorithm

Transportation Routines BA Goal:  workplace Trip segments ( Including transportation modes )  Home to Bus stop A on Foot  Bus stop A to Bus stop B on Bus  Bus stop B to workplace on Foot Location and velocity Workplace

Dynamic Bayesian Network x k-1 z k-1 zkzk xkxk Time k-1 Time k x= GPS reading z t k-1 tktk g k-1 gkgk Goal g Trip segment t

Rao-Blackwellised Particle Filtering Inference: estimate the distribution of current state given the measurements up to now Particle filters:  a Monte Carlo method capable of doing inference in a DBN  Approximate the distribution of the state using a number of weighted samples (particles) Rao-Blackwellised particle filters  One variant of particle filter that combines the Monte Carlo approximation with exact inference techniques

Learning Encoded in the DBN a lot of interesting information Learning the parameters of the DBN equates to learning the information about the user  Usual goals  Routine parking spots and bus stops  Most likely trajectories getting to the goals  ….

Expectation-Maximization Algorithm Unsupervised learning: we don’t need to label the data manually! Capable of learning all the parameters simultaneously A Monte Carlo version of EM is used for efficiency

Learning Goals

Learning Trajectories Going to the workplaceGoing home Zoom into the area near the workplace

Error Detection Track using two different trackers simultaneously  The first tracker assumes normal activity while the second tracker assumes abnormal activity  At each time, estimate the probability for each tracker

Thanks for your attention!