Archives—March 2012

CS, Center for Computer Systems Research and ECE Seminar – Brian VanVoorst, Speaker

Title: An Introduction to Point Cloud Understanding
Brian VanVoorst, MTU Alumni & Technical Director of BBN Technologies

Thursday, March 22,2012 – 135 Fisher Hall – 2:00 PM

Abstract: A point cloud is a collection of 3D points from a 3D sensor such as a LIDAR, stereo camera, or a Microsoft Kinect system. These 3D sensors are used in applications of robotics, mapping (such as the Google Street View platforms), and entertainment. At BBN there are multiple projects under way with a common theme of “point cloud understanding.” Point cloud understanding is an area of computer vision research in which algorithms are developed to extract knowledge from point clouds. In this talk an overview of 3D sensors and their point clouds, discuss challenges computer scientists face in processing point clouds, explain some of the key algorithms and data structures, highlight the differences between point cloud understanding and image understanding, and explore opportunities for sensor fusion. I will draw heavily upon the real-world challenges we face in our ongoing research projects. This talk will be accessible to computer scientists and engineers at all levels.

Biography: Brian VanVoorst joined BBN Technologies in 2008 as a Technical Director to help form the BBN Technologies office in Minnesota. He has more than 19 years of experience working on and leading research and development programs. His most recent work is in the area of the automated understanding of LIDAR point clouds. His previous work has been in many areas, including real-time and fault-tolerant systems, mobile ad-hoc networking, parallel processing, and parallel system benchmarking. He also has worked extensively with robotics and was part of a team that was a finalist for the DARPA Urban Challenge. Before coming to BBN, VanVoorst was a researcher at Honeywell Labs for 14 years and spent two years at the NASA Ames Research Center. VanVoorst earned his bachelor’s and master’s degrees in computer science from Michigan Technological University. From 1999–2001 he held a lectureship position at Michigan Tech and taught in the Computer Science Department while continuing to work for Honeywell. He holds one patent with four applications pending and has published more than 20 papers in conference proceedings and journals.


5th Annual BonzAI Brawl Programming Contest

Put your AI to the test and conquer the nine realms! On March 31, 2012 the 5th Annual BonzAI Brawl programming competition will take place in the CS department at Michigan Technological University. The programming will be an all day event, where teams of 1 to 3 contestants will implement an AI player for a game. The contestants will be given the details of the API the day of the competition and must design a winning strategy within the 8 hours allotted. After coding ends, the AIs are pitted against each other, in a tournament (known as the BRAWL). Spectators are welcome to attend and cheer for their favorite AI at the BRAWL. For more information about BonzAI Brawl or to register your team, visit http://wics.students.mtu.edu/bonzai. All teams must register by March 23, 2011.

Sponsored in part by a donation from LaSalleTech, Consistacom, Jackson, and the CS Department.


CS Department Seminar – Dr. Timothy Havens, Speaker

Department of Computer Science Seminar
February 27, 2012 – 4:04 PM – Room G005 – Rekhi Hall
Title: “Fuzzy Kernel Clustering of Large Scale Biomedical and Bioinformatics Data”

Dr. Timothy Havens

Abstract:
Since the early 1990’s, the ubiquity of personal computing technology has produced an abundance of staggeringly large data sets—it is estimated that Facebook alone logs over 25 terabytes of data per day and large bioinformatics data sets that integrate microarrays, sequences, and ontology annotations continue to grow. To compound this fact, these data sets are populated from disparate, often unknown, sources and are in a wide-range of formats. There is a great need for systems by which one can elucidate the similarity among and between groups in these data sets and produce easy-to-understand visualizations of the results. In this talk, I will discuss a method for efficiently and accurately approximating the solution of the kernel c-means clustering algorithm, specifically focusing on the fuzzy variant. Kernel clustering has been shown to be effective for data sets where the groups are not linearly separable in the input space or are high-dimensional. However, kernel fuzzy c-means (kFCM) presents computation and storage requirement challenges: clustering 500,000 objects requires 1 terabyte of main memory. I will show that on medium scale data (~50,000 objects) the approximate kFCM (akFCM) algorithm gives up to three orders of magnitude speed-up and a constant factor reduction in memory footprint with little-to-no degradation in performance, as compared to literal kFCM. I also demonstrate that akFCM performs well on large-scale data (>500,000 objects), including magnetic resonance imaging volumes. Last, I will apply the clustering method to bioinformatics data composed of genes described by Gene Ontology annotations to show how akFCM can be used for comparative genomics.


Bo Yu Selected to Receives Michigan Tech SURF Award

The Michigan Technological University Department of Computer Science is proud to announce that Bo Yu, a senior in the CS, as been selected to receive the Michigan Tech SURF (Summer Undergraduate Research Fellowship) award. Bo has received the maximum award given for this fellowship – $3300.

Bo’s research, under the mentoring of Dr. Ali Ebnenasir, is titled, “Towards Designing a Fault-Tolerant Scheduler for the OkL4 Microkernal.” OkL4 is a small microkernel found within millions of smart phones. The research involves studying the task queue of the OkL4 scheduler, analyzing the impact of transient faults on the task queue, designing recovery from transient faults, and refining recovery back to the level of the OkL4 source code.

Congratulations to Bo and Dr. Ali Ebnenasir!