| Accepted Videos | ||||||
|---|---|---|---|---|---|---|
| # | Title | Developer(s) | Affiliation(s) | Length | Student | |
| 1 | Galactic Arms Race (GAR): Automatic Content Generation In a Multiplayer Online Video Game | Erin J. Hastings, Ratan K. Guha, and Kenneth O. Stanley | University of Central Florida | 2:20 | Yes | |
| Description: This video showcases a new AI algorithm called cgNEAT that automatically generates content in video games. To demonstrate this new technique, we created a near-commercial-quality game called Galactic Arms Race (GAR) in which the weapons systems are entirely invented by the game itself. The cgNEAT method, which is short for content-generating NeuroEvolution of Augmenting Topologies, evolves new weapons (which are controlled by artificial neural networks) by varying the most popular weapons of the past. In this way, an evolutionary process causes the algorithm to explore the space of weapons as the game is played, producing a never-ending supply of novel and functional content. The aim is to show that AI can be sophisticated enough to produce some of the content in games without the need for artists or programmers, by observing what players liked in the past. The video presents a montage of actual gameplay that demonstrates the surprising variety of compelling weapons invented by the game itself. It also explicates the underlying AI technology more through action than through words. For more information on GAR, which will be released this Spring, please visit: http://gar.eecs.ucf.edu. | ||||||
| 2 | EDGE: Enhanced Device for Geospatial Exploration | Feras Batarseh | University of Central Florida | 4:13 | Yes | |
| Description: This video is a presentation of a robotics simulation project. Microsoft Robotics Project was used for the development of this project. The video shows the robot/agent in the simulation environment, looking for hostages, takes pictures of the hostage and locates their positions. the video show how the robot scans the area, and avoids walls using its laser. The video shows the screens of the software, with a nice rock song! and a brief description of the methods, and algorithms used by the robot. | ||||||
| 3 | Learning Kinematic Models of Articulated Objects | Juergen Sturm, Vijay Pradeep, Cyril Stachniss, Christian Plagemann, Kurt Konolige, and Wolfram Burgard | University of Freiburg | 1:00 | Yes | |
| Description: Robots operating in home environments must be able to interact with articulated objects such as doors or drawers. Ideally, robots are able to autonomously infer articulation models by observation. In this video, we briefly present an approach for learning kinematic models by inferring the connectivity of rigid parts and the articulation models for the corresponding links. Our method uses a mixture of parameterized and parameter-free (Gaussian process) representations and finds low-dimensional manifolds that provide the best explanation of the given observations. Our approach has been implemented and evaluated using real data obtained in various realistic home environment settings. Corresponding paper: http://www.informatik.uni-freiburg.de/~sturm/media/sturm09ijcai.pdf | ||||||
| 4 | The Autonomous City Explorer | Andrea Bauer, Klaas Klasing, Tingting Xu, Stefan Sosnowski, Georgios Lidoris, Quirin Mühlbauer, Tianguang Zhang, Florian Rohrmüller, Dirk Wollherr, Kolja Kühnlenz, and Martin Buss | Technische Universität München | 4:36 | Yes | |
| Description: This video presents the Autonomous City Explorer (ACE) project. Its goal was to create a robot capable of navigating in an unknown urban environments without the use of prior map knowledge or GPS data. The robot had to find its way solely by interacting with pedestrians and building a topological representation of its surroundings. This video outlines the necessary ingredients for successful low-level navigation on sidewalks, vision processing in cluttered outdoor environments, and information retrieval from pedestrians. More information can be found at www.ace-robot.de. | ||||||
| Nominations: Best, Most Innovative, Best Sound Track | ||||||
| 5 | Applying Case-Based Reasoning to Texas Hold'em Poker | Jonathan Rubin and Ian Watson | University of Auckland | 4:39 | Yes | |
| Description: This video summarizes the use of Case-Based Reasoning principles to the area of of a Texas Hold'em poker. Introductions to CBR and Texas Hold'em are included as well as a description about two poker-bots which we have developed that make use of CBR to play poker. Results and conclusions of our research are briefly presented. For more information you can visit our website: http://www.cs.auckland.ac.nz/research/gameai. | ||||||
| 7 | Motion Synthesis and Control Learning for (Un)Knotting Deformable Linear Objects | Sandhya Prabhakaran | University of Edinburgh and University of Basel | 4:56 | Yes | |
| Description: This research deals with motion planning and control for Deformable Linear Objects (DLOs). It is still a complex task to get a robot manipulate a rope or cloth. To realise this vision of getting a robot to handle dexterous objects, we have taken the simplest object i.e. a DLO for our study purpose. The operations performed on the DLO are knot-(un)tying. The DLO is thus parameterised as a Knot and we make the DLO (un)tied into various knot types. The mathematical branch of Knot Theory is extensively used here to realise the Knots. The configuration space that the Knot can move is computed by the Knot Energy. We use the Minimum Distance Knot energy here. With this, we create a hierarchical graph structure with nodes corresponding to optimal knot configurations obtained by optimising this Knot energy functional. Thus by navigating this graph, we are able to (un)tie various knots. The study looks into 3 simple and 2 complex knot types. Motion control while (un)tying is also brought about using the SARSA [Reinforcement Learning] algorithm. The motion planner is resilient to perturbations as well. Thus by devising Knot Energy together with SARSA, we have built a multi-scale, reactive knot (un)tying motion planner. Results show that our method is incredibly faster than normal Probabilistic and feedback control methods. For more, please refer to my MSc thesis titled 'Multi-scale, Reactive Motion Planning with Deformable Linear Objects' at http://www.inf.ed.ac.uk/publications/thesis/online/IM080596.pdf or for the complete version including Motion control at http://www.mediafire.com/file/dozd2zm3mam/MSc thesis - SARSA and Knot energy version.pdf. | ||||||
| 8 | AIspace: Tools for Learning Artificial Intelligence | Byron Knoll, Alan Mackworth, David Poole, Giuseppe Carenini, Jacek Kisyński, Cristina Conati, and Holger Hoos | University of British Columbia | 1:00 | Yes | |
| Description: This video summarizes our work on a set of interactive algorithm visualization tools for teaching and learning AI. The tools cover many of the topics that would be in an intro AI course. They are freely available online at http://aispace.org. They were developed at the Laboratory for Computational Intelligence at the University of British Columbia. | ||||||
| Nominations: Best Short | ||||||
| 9 | The Dinochrome Brigade | William M. Spears, and Paul M. Maxim | University of Wyoming | 5:00 | No | |
| Description: This video provides a short introduction to our technology for robot localization, and then shows how this technology enables various robotic tasks. Brief demonstrations are shown for moving robot formations, uniform coverage of a region, chain formations, and collaborative pulling. The video is meant to highlight our accomplishments, but is not a tutorial. More information can be found at: http://www.cs.uwyo.edu/~wspears/maxelbot and http://www.cs.uwyo.edu/~wspears/pubs.html. | ||||||
| 10 | Multi-Camera People Tracking Presented by a Humanoid | Kyungnam (Ken) Kim | HRL Laboratories, USA | 4:59 | No | |
| Description: A humanoid gives a talk, like a human presenter, about a real-time vision system using multiple cameras for people tracking in a room. The humanoid demonstrates a few perception/control capabilities including visually guided hand/arm control. It briefly describes the algorithm used and shows some people tracking results including each person's trajectory and walking direction, and reports the number of people in the room. The vision system with two quad-core 3.0 Ghz CPUs runs in real-time. In the end of the video, a game application based on person tracking is presented. | ||||||
| 11 | Copycat Hand for All | Motomasa Tomida, Takanobu Tanimoto, Kiyoshi Hoshino | University of Tsukuba | 2:00 | Yes | |
| Description: "Copycat Hand for All" is a robot system that imitates the human motions, by visually estimating the human hand and arm postures at a high speed and with high accuracy. The system uses only one high-speed camera and note PC. But once you stand in front of the robot and move your hand and arm freely, it reproduces your behavior without time delay. At the first stage, our system uses coarse screening by the proportional information on the hand images which roughly correspond to forearm rotation, bending of the thumb or four fingers. And then, at the second stage, it performs a detailed search for the selected candidates. The estimation error is less than one degree in the joint angle, and the processing time is 80 fps or more. We are now enhancing flexibility for those having thick/thin, long/short, or deeply hooking fingers, through using different typical hand CG images featuring different bone length and thickness and joint movable ranges in order to permit the system to respond accurately to people of all ages and both sexes, including foreign people. For more information, please see our lab's page at http://www.kz.tsukuba.ac.jp/~hoshino/AI/copycat.pdf. | ||||||
| Nominations: Best Student | ||||||
| 13 | Toward Interactive Learning of Container and Non-Container Objects | Shane Griffith, Jivko Sinapov, Matthew Miller, and Alexander Stoytchev | Iowa State University | 1:55 | Yes | |
| Description: This video highlights our work toward creating robots that interactively learn about container and non-container objects. It demonstrates that a robot can distinguish between them by dropping a block in the area above the object and observing the co-movement patterns while pushing the object. Also, the video illustrates that the robot can learn a perceptual model of containers to identify novel containers in the environment. Check out the lab's webpage at http://www.ece.iastate.edu/~alexs/lab/ or the first author's webpage at http://www.ece.iastate.edu/~shaneg/research.html for more information. | ||||||
| Nominations: Best Narration | ||||||
| 14 | A Plan-Based Machinima Creation System | Mike Dominguez | North Carolina State University | 4:51 | Yes | |
| Description: This video describes our plan-based machinima creation system and concludes with an example film created with it. | ||||||
| 15 | How to Cook as Perfect Love Story | Amélie Cordier | Lyon 1 University/td> | 4:32 | No | |
| Description: This movie is about the Taaable project (http://taaable.fr), a Web-based CBR Cooking system (also a participant to the Computer Cooking Contest). The movie illustrates on a real situation how a CBR system can help a human to solve a cooking problem and how it learns in case of failure. The main topics addressed by the movie are: case-based reasoning and interactive learning. | ||||||
| Nominations: Best, Best Sound Track | ||||||
| 16 | CogSketch: Open-Domain Sketch Understanding for Research and Education | Kate Lockwood, Andrew Lovett, Jeffrey Usher, Penny Yin, Jon Wetzel, and Kenneth Forbus | Northwestern University | 4:59 | Yes | |
| Description: This video describes CogSketch, a sketch understanding system being developed for research and educational purposes. CogSketch automatically generates symbolic representations of sketches. These representations can be used as the input to reasoning systems. The video demonstrates two educational applications of CogSketch: Worksheets, which can provide students with automatic feedback by comparing their sketches to teacher sketches and identifying differences; and the Design Buddy, which uses qualitative mechanics to analyze engineering design sketches. | ||||||
| 17 | Bodies In Motion: Dynamic Motion Capture | Daniel Byers, Marek Vondrak, Leonid Sigal, and Odest Chadwicke Jenkins | Brown University and University of Toronto | 6:15 (will edit) | Yes | |
| Description: We explore the use of full-body 3D physical simulation for human kinematic tracking from monocular and multi-view video sequences within the Bayesian filtering framework. Towards greater physical plausibility, we consider a human's motion to be generated by a "feedback control loop", where Newtonian physics approximates the rigid-body motion dynamics of the human and the environment through the application and integration of forces. The result is more faithful modeling of human-environment interactions, such as ground contacts, resulting from collisions and the human's motor control. For more information, please see: http://robotics.cs.brown.edu/projects/dynamical_tracking/ | ||||||
| Nominations: Best, Most Innovative | ||||||
| 18 | Situated Interaction | Dan Bohus and Eric Horvitz | Microsoft Research | 4:11 | No | |
| Description: This video provides an overview of the Situated Interaction project at Microsoft Research, an effort towards developing open-world interactive systems that can embed interaction and computation deeply into the natural flow of daily tasks, activities and collaborations. The video highlights project goals and reviews a prototype platform which weaves together several component technologies, including learning and inference about activities and goals, speech recognition and synthesis, vision, conversational scene analysis, and multiparty dialog management, to support fluid interactions with multiple users in an open-world context. The video illustrates several challenges and competencies with an application running on the platform, named Receptionist, that handles problems in the domain of building receptionists. More information about the project and related research are available at http://research.microsoft.com/en-us/um/people/dbohus/research_situated_interaction.html. | ||||||
| Nominations: Most Informative | ||||||
| 21 | Reinforcement Learning by Example | Cosmin Paduraru, Robert West, and Imad Khoury | McGill University | 4:59 | Yes | |
| Description: A brief introduction to reinforcement learning, using the task of bartending as an example. | ||||||
| Nominations: Best Educational | ||||||
| 22 | Real Live Robot Learning | Michael Littman and Kaushik Subramanian | Rutgers University | 3:41 | No | |
| Description: We created a reinforcement-learning demo---a simple robot navigation task---and took it to the public to teach them about AI and robotics. The video shows the system adapting in real time to various modifications to the robot's design and provides a very gentle introduction to the idea of model-based reinforcement learning. | ||||||
| Nominations: Best Educational, Best Narration | ||||||
| 23 | Using Entropy to Distinguish Shape versus Text in Hand-Drawn Diagrams | Akshay Bhat and Tracy Hammond | Texas A&M University, College Station | 1:00 | Yes | |
| Description: Most sketch recognition systems are accurate in recognizing either text or shape (graphic) ink strokes,but not both. Distinguishing between shape and text strokes is, therefore, a critical task in recognizing hand-drawn digital ink diagrams that contain text labels and annotations. We have found the entropy rate to be an accurate criterion of classification. We found that the entropy rate is significantly higher for text strokes compared to shape strokes and can serve as a distinguishing factor between the two. The paper has been accepted for publication in the 2009 IJCAI proceedings. This video shows our system in action. Please visit our lab here: http://srlweb.cs.tamu.edu/srlng/home. | ||||||
| 24 | News at Seven: The Future of the Future | Lisa Gandy, Nathan D. Nichols, and Kristian J. Hammond | Northwestern University | 1:00 | Yes | |
| Description: This video summarizes News at Seven, an automated news system. The system is able to find relevant text, process that text, and supplement it with images, video, and blogger responses. The final output of the system is an online Flash presentation that uses animated avatars with generated speech and is modeled after traditional nightly news broadcast. Current segments include a Movie Review, an Entertainment Segment, and a Louis Black style rant segment. More information about News at Seven can be found at http://newsatseven.com/. News at Seven is also currently live at http://www.zap2it.com/news/zap-news-at-7,0,6717570.htmlstory. | ||||||
| Nominations: Best Short, Best Student | ||||||
| 25 | Penso BCI System | Dorian Peters, Rafael A. Calvo, and Payam Aghaeipour | University of Sydney | 2:11 | No | |
| Description: This video reveals how Brain Computer Interfaces work in basic terms by demonstrating a BCI system called Penso. By providing a less formal nutshell view of the basic principles of a BCI system in a language accessible to a general audience, this 2-minute video has sparked interest and enthusiasm in prospective postgraduate students in the BCI area. For more information, please see our team's page at http://www.weg.ee.usyd.edu.au/projects/penso. | ||||||
| 26 | Write. Reflect. Polish. | Dorian Peters and Rafael A. Calvo | University of Sydney | 2:53 | No | |
| Description: This video introduces a writing support too called "Glosser" that leverages language processing and text analysis techniques to analyze and provide feedback to students as they write an essay. Although the video was designed to welcome engineering students to a tool they would be using in their first year, it also ignited student interest in natural language processing and related AI technologies, by showing how certain technologies can be applied in the real world. For more information, please see our team's page at http://www.weg.ee.usyd.edu.au/projects/glosser. | ||||||
| 27 | Robots to the Rescue: Mixed-initiative human-robot teaming for disaster response | Cynthia Breazeal, Matt Berlin, Paula Aguilera, Kenton Williams, Julie A. Adams, Philipp Robbel, Sanford Freedman, Jonathan How, Aditya Undurti, Jesse Gray, Stefanie Tellex, Nicholas Roy, Tom Kollar, Sigurdur Orn Adalgeirsson, Jason Alonso, Fardad Faridi, Jun Ki Lee, Mikey Siegel, Sophie Wang, and Jonathan Williams | MIT and Vanderbilt University | 5:00 | No | |
| Description: The humanoid robot Nexi and a team of robot helicopters are deployed in response to a simulated fire aboard a Navy ship. Working together with a remote human operator, the robots search for survivors and guide them to safety. A mixed-initiative tasking system allows the human operator to specify team goals via a tasking interface, and also allows goals to be generated by the robots as a result of local observations and interactions with victims. The robots autonomously handle the details of navigation and task execution, and communicate their location, task status, and important observations to the operator via the interface. The interface also allows the robots to ask for the operator's help with difficult recognition problems, such as confirming the location of a victim from a partial identification. | ||||||
| Nominations: Best, Best Sound Track | ||||||
| 28 | Improving Offensive Performance Through Opponent Modeling | Kennard R. Laviers and Gita Sukthankar | University of Central Florida | 4:48 | Yes | |
| Description: Although in theory opponent modeling can be useful in any adversarial domain, in practice it is both difficult to do accurately and to use effectively to improve game play. In this video, we present an approach for online opponent modeling and illustrate how it can be used to improve offensive performance in the Rush 2008 football game. In football, team behaviors have an observable spatio-temporal structure, defined by the relative physical positions of team members over time; we demonstrate that this structure can be exploited to recognize football plays at a very early stage of the play using a supervised learning method. Based on the teams' play history, our system evaluates the competitive advantage of executing a play switch based on the potential of other plays to increase the yardage gained and the similarity of the candidate plays to the current play. In this video, we investigate two types of play switches: 1) whole team and 2) subgroup. Both types of play switches improve offensive performance, but modifying the behavior of only a key subgroup of offensive players yields greater improvements in yardage gained. | ||||||
| Nominations: Best Student | ||||||
| 29 | WiiGesture | Michael Delp | University of Alberta | 1:00 | Yes | |
| Description: WiiGesture is a gesture recognition program for actions that use accelerometer data. It uses artificial intelligence to classify gestures using a wiimote from a few examples of each gesture. This was a project for a Machine Learning class. Many algorithms were tried, like LCSS, Bagged Trees, SVMs, and Fast Fourier Transforms and the video highlights the one we found most useful (Cross Correlation). This was a really fun project that has real world applications in the video game industry, and I hope it encourages students to consider studying Artificial Intelligence. | ||||||
| Nominations: Best Short | ||||||
| 30 | Real-Life Reinforcement Learning | Ali Nouri and Michael L. Littman | Rutgers University | 2:49 | Yes | |
| Description: A lot of the research in the field of reinforcement learning has been focused/tested on simulation domains. Some of the characteristics of real-life domains, such as the shape of their noise function or imperfect perception that results in violation of Markovianness are typically neglected in the simulation. In this video, we promote performing reinforcement learning research on real problems to make sure algorithms are more robust to these design imperfections. | ||||||
| Nominations: Most Informative | ||||||
| 31 | Playbook: a new approach to tasking interfaces | David Musliner and Bryan Bell | Smart Information Flow Technologies | 4:29 | No | |
| Description: Playbook is a system developed at Smart Information Flow Technologies that allows the delegation of complex tasks from humans to multiple unmanned systems. Through this method, an operator can declare high level instructions that are understandable by unmanned aerial vehicles, allowing them to automatically calculate most efficient means to completing their goal, dramatically simplifying the operator's workload. More information at www.sift.info. | ||||||
| 33 | Conversational Virtual Role Players | Joel Harris and Prasan Samtani | Alelo, Inc. | 3:53 | No | |
| Description: This video showcases a technology called Virtual Role Players (VRP) for implementing conversational virtual humans that engage in spoken dialog, and exhibit culturally appropriate behavior. It is designed for use as a plug-in to training simulation systems. The version displayed is integrated with the Virtual Battlespace 2 (VBS2) mission rehearsal environment. | ||||||
| 34 | The Stanford Autonomous Helicopter | Adam Coates, Pieter Abbeel, and Andrew Y. Ng | Stanford University | 4:40 | Yes | |
| Description: Stanford's Autonomous Helicopter project pushes the limits of autonomous flight control by teaching a computer to fly a competition-class remote controlled (RC) helicopter through a range of aerobatic stunts. Our apprenticeship learning approach learns to fly the helicopter by observing human demonstrations and is capable of a wide variety of expert maneuvers. In many cases, it can even exceed the performance of the human expert from which it learned. http://heli.stanford.edu. | ||||||
| Nominations: Best Student, Most Innovative | ||||||
| 35 | The Intelligent "Dynamic" Workbook for Learning Written East Asian Languages | Paul Taele and Tracy Hammond | Texas A&M University | 0:54 | Yes | |
| Description: Our video summarizes one of our recent projects regarding sketch recognition for the domain of written East Asian languages such as Chinese and Japanese. Our research focuses on developing computer-assisted language instruction (CALI) systems which provide human instructor-level assessment of students' written East Asian writings, both for visual structure and written technique. That is, we strive to provide an intelligent "dynamic" workbook for supplementing current East Asian language programs, in order to alleviate the difficulties of learning written East Asian. | ||||||
| 36 | Little Robot Goes Missing | Eszter Szepesvari, Reka Szepesvari, and Csaba Szepesvari | Vernon Barford Junior High School (Edmonton), Harry Ainlay High School (Edmonton), and the University of Alberta | 2:54 | Yes | |
| Description: In this video we demonstrate how robots are able to find their location on a map. The video is based on the project work of 4th year students of the University of Alberta. | ||||||
| 37 | Robotic Secrets Revealed, Episode 001 | Anthony M. Harrison, Benjamin R. Fransen, Magdalena Bugajska, and J. Gregory Trafton | Naval Research Laboratory | 4:40 | No | |
| Description: Using Three-Cup Shuffle Magic Trick as a backdrop, we present a system which performs head, body pose, and fiducial-based object tracking all in 3D as well as hand and head gesture recognition and production. The sensor born information and the capabilities are tightly integrated in a psychologically plausible manner within an embodied version of the ACT-R cognitive architecture, ACT-R/E. The Mobile-Dextrous-Social (MDS) Robot interprets and uses gestures, including deictic and symbolic hand gestures as well as head nods and shakes. The overall approach provides us with a powerful, integrated system facilitating many different avenues of research in human-robot interaction. | ||||||
| Nominations: Best Educational, Most Informative | ||||||
| 39 | Casey's Quest: Transfer Learning for Adversarial Environments | Philip Moore, Matthew Molineaux, and Kalyan Gupta | Knexus Research Corporation | 4:59 | No | |
| Description: A dramatization of Transfer Learning research through the journey of an 8-bit football player. Based on research conducted by David Aha, Matthew Molineaux, and Gita Sukthankar for ICCBR'09. More information is available at http://www.knexusresearch.com/projects/rush. | ||||||
| Nominations: Best, Best Narration, Best Sound Track | ||||||