Disney Research Pittsburgh
Disney Research Pittsburgh
Yamada M.,Yahoo! |
Jitkrittum W.,University College London |
Sigal L.,Disney Research Pittsburgh |
Xing E.P.,Carnegie Mellon University |
Sugiyama M.,Tokyo Institute of Technology
Neural Computation | Year: 2014
The goal of supervised feature selection is to find a subset of input features that are responsible for predicting output values. The least absolute shrinkage and selection operator (Lasso) allows computationally efficient feature selection based on linear dependency between input features and output values. In this letter, we consider a feature-wise kernelized Lasso for capturing nonlinear input-output dependency.We first show that with particular choices of kernel functions, nonredundant features with strong statistical dependence on output values can be found in terms of kernel-based independence measures such as the Hilbert- Schmidt independence criterion.We then show that the globally optimal solution can be efficiently computed; thismakes the approach scalable to high-dimensional problems. The effectiveness of the proposed method is demonstrated through feature selection experiments for classification and regression with thousands of features. © 2013 Massachusetts Institute of Technology.
Mistry M.,Disney Research Pittsburgh |
Righetti L.,University of Southern California
Robotics: Science and Systems | Year: 2012
The operational space formulation (Khatib, 1987), applied to rigid-body manipulators, describes how to decouple task-space and null space dynamics, and write control equations that correspond only to forces at the end-effector or, alternatively, only to motion within the null space. We would like to apply this useful theory to modern humanoids and other legged systems, for manipulation or similar tasks, however these systems present additional challenges due to their underactuated floating bases and contact states that can dynamically change. In recent work, Sentis et al. derived controllers for such systems by implementing a task Jacobian projected into a space consistent with the supporting constraints and underactuation (the so called support consistent reduced Jacobian). Here, we take a new approach to derive operational space controllers for constrained underactuated systems, by first considering the operational space dynamics within projected inverse-dynamics (Aghili, 2005), and subsequently resolving underactuation through the addition of dynamically consistent control torques. Doing so results in a simplified control solution compared with previous results, and importantly yields several new insights into the underlying problem of operational space control in constrained environments: 1) Underactuated systems, such as humanoid robots, cannot in general completely decouple task and null space dynamics. However, 2) there may exist an infinite number of control solutions to realize desired task-space dynamics, and 3) these solutions involve the addition of dynamically consistent null space motion or constraint forces (or combinations of both). In light of these findings, we present several possible control solutions, with varying optimization criteria, and highlight some of their practical consequences.
Mistry M.,Disney Research Pittsburgh |
Buchli J.,University of Southern California |
Schaal S.,University of Southern California
Proceedings - IEEE International Conference on Robotics and Automation | Year: 2010
Model-based control methods can be used to enable fast, dexterous, and compliant motion of robots without sacrificing control accuracy. However, implementing such techniques on floating base robots, e.g., humanoids and legged systems, is non-trivial due to under-actuation, dynamically changing constraints from the environment, and potentially closed loop kinematics. In this paper, we show how to compute the analytically correct inverse dynamics torques for model-based control of sufficiently constrained floating base rigid-body systems, such as humanoid robots with one or two feet in contact with the environment. While our previous inverse dynamics approach relied on an estimation of contact forces to compute an approximate inverse dynamics solution, here we present an analytically correct solution by using an orthogonal decomposition to project the robot dynamics onto a reduced dimensional space, independent of contact forces. We demonstrate the feasibility and robustness of our approach on a simulated floating base bipedal humanoid robot and an actual robot dog locomoting over rough terrain. ©2010 IEEE.
Liu J.,Pennsylvania State University |
Carr P.,Disney Research Pittsburgh |
Collins R.T.,Pennsylvania State University |
Liu Y.,Pennsylvania State University
Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition | Year: 2013
We employ hierarchical data association to track players in team sports. Player movements are often complex and highly correlated with both nearby and distant players. A single model would require many degrees of freedom to represent the full motion diversity and could be difficult to use in practice. Instead, we introduce a set of Game Context Features extracted from noisy detections to describe the current state of the match, such as how the players are spatially distributed. Our assumption is that players react to the current situation in only a finite number of ways. As a result, we are able to select an appropriate simplified affinity model for each player and time instant using a random decision forest based on current track and game context features. Our context-conditioned motion models implicitly incorporate complex inter-object correlations while remaining tractable. We demonstrate significant performance improvements over existing multi-target tracking algorithms on basketball and field hockey sequences several minutes in duration and containing 10 and 20 players respectively. © 2013 IEEE.
Lan T.,Simon Fraser University |
Sigal L.,Disney Research Pittsburgh |
Mori G.,Simon Fraser University
Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition | Year: 2012
We present a hierarchical model for human activity recognition in entire multi-person scenes. Our model describes human behaviour at multiple levels of detail, ranging from low-level actions through to high-level events. We also include a model of social roles, the expected behaviours of certain people, or groups of people, in a scene. The hierarchical model includes these varied representations, and various forms of interactions between people present in a scene. The model is trained in a discriminative max-margin framework. Experimental results demonstrate that this model can improve performance at all considered levels of detail, on two challenging datasets. © 2012 IEEE.
Israr A.,Disney Research Pittsburgh |
Poupyrev I.,Disney Research Pittsburgh
Conference on Human Factors in Computing Systems - Proceedings | Year: 2011
Tactile Brush is an algorithm that produces smooth, two-dimensional tactile moving strokes with varying frequency, intensity, velocity and direction of motion. The design of the algorithm is derived from the results of psychophysical investigations of two tactile illusions - apparent tactile motion and phantom sensations. Combined together they allow for the design of high-density two-dimensional tactile displays using sparse vibrotactile arrays. In a series of experiments and evaluations we demonstrate that Tactile Brush is robust and can reliably generate a wide variety of moving tactile sensations for a broad range of applications. Copyright 2011 ACM.
Tena J.R.,Disney Research Pittsburgh |
De La Torre F.,Carnegie Mellon University |
Matthews I.,Disney Research Pittsburgh
ACM Transactions on Graphics | Year: 2011
Linear models, particularly those based on principal component analysis (PCA), have been used successfully on a broad range of human face-related applications. Although PCA models achieve high compression, they have not been widely used for animation in a production environment because their bases lack a semantic interpretation. Their parameters are not an intuitive set for animators to work with. In this paper we present a linear face modelling approach that generalises to unseen data better than the traditional holistic approach while also allowing click-and-drag interaction for animation. Our model is composed of a collection of PCA sub-models that are independently trained but share boundaries. Boundary consistency and user-given constraints are enforced in a soft least mean squares sense to give flexibility to the model while maintaining coherence. Our results show that the region-based model generalizes better than its holistic counterpart when describing previously unseen motion capture data from multiple subjects. The decomposition of the face into several regions, which we determine automatically from training data, gives the user localised manipulation control. This feature allows to use the model for face posing and animation in an intuitive style. © 2011 ACM.
Zheng Y.,Disney Research Pittsburgh |
Yamane K.,Disney Research Pittsburgh
IEEE Transactions on Automation Science and Engineering | Year: 2013
Ray shooting is a well-studied problem in computer graphics and also has applications in robotics such as collision detection and contact force optimization. Unfortunately, most ray-shooting algorithms developed for graphics applications only allow 3-dimensional (3-D) objects represented as triangle meshes, and therefore are not suited for objects with parametric surfaces or general convex sets in high-dimensional space which robotics applications often require. In contact force optimization, for example, the problem is in the 6-dimensional (6-D) wrench space and it is desirable to consider the nonlinear friction cone without approximating it by a pyramid. This paper discusses existing and novel geometry-based ray-shooting algorithms applicable to general convex sets, and compares their performances in two robotics applications: computing the distance between two convex objects and optimizing contact forces in grasping. © 2004-2012 IEEE.
Ishiguro Y.,Disney Research Pittsburgh |
Poupyrev I.,Disney Research Pittsburgh
Conference on Human Factors in Computing Systems - Proceedings | Year: 2014
We propose technology for designing and manufacturing interactive 3D printed speakers. With the proposed technology, sound reproduction can easily be integrated into various objects at the design stage and little assembly is required. The speaker can take the shape of anything from an abstract spiral to a rubber duck, opening new opportunities in product design. Furthermore, both audible sound and inaudible ultrasound can be produced with the same design, allowing for identifying and tracking 3D printed objects in space using common integrated microphones. The design of 3D printed speakers is based on electrostatic loudspeaker technology first explored in the early 1930s but not broadly applied until now. These speakers are simpler than common electromagnetic speakers, while allowing for sound reproduction at 60 dB levels with arbitrary directivity ranging from focused to omnidirectional. Our research of 3D printed speakers contributes to the growing body of work exploring functional 3D printing in interactive applications.
News Article | October 28, 2016
Brandon Lucia, an assistant professor of electrical and computer engineering at Carnegie Mellon University, and his Ph.D. student Alexei Colin created the first programming language designed to build reliable software for intermittent, energy-harvesting computers. Colin will present the work at the 2016 SPLASH conference in Amsterdam, Netherlands, on November 3rd. "Energy is not always available in the environment for a device to harvest," explains Lucia. "Intermittent operation makes it difficult to build applications because existing software programming languages—and programmers themselves—assume that energy is a continuously available resource." The innovative new programming language, called Chain, asks an application developer to define a set of computational tasks that compute and exchange data through a novel way of manipulating the computer's memory, called a channel. Chain guarantees that tasks execute correctly despite arbitrary power failures. "When power is not continuously available, power failures disrupt the software's execution, often leading to unrecoverable errors," says Lucia. "Chain solves this problem by requiring computational tasks in the program to use a novel channel-based memory abstraction that ensures tasks complete without error." Channel-based memory is the key to Chain's ability to avoid software errors: regardless of when power fails, channels ensure that a computational task always has an intact version of the data it needs when power resumes. Restarting a Chain program after a failure has virtually zero time cost because Chain does not rely on an expensive, conventional approach, like memory checkpointing. The extreme scarcity of energy makes efficient restarting essential for energy-harvesting applications including IoT devices and implantable or ingestible medical devices. "Chain provides important reliability guarantees in a familiar and flexible programming interface that is well-positioned to be the foundation for today's and future energy-harvesting applications," says Lucia. Lucia, Colin, and Dr. Alanson Sample, a collaborator at Disney Research Pittsburgh, worked together to push Chain into real-world deployment; early next year, in cooperation with nano-satellite company KickSat, software written in Chain will run on-board two tiny, postage stamp-sized satellites in a low-earth orbit of Earth. Once in orbit, these satellites will use tiny solar panels to harvest solar energy, powering them to collect and process sensor data and send information back to earth. While satellites are typically powered by solar energy, these satellites will be the first with the strong software correctness guarantees furnished by Chain ensuring continuous, reliable operation. "The potential benefit of reliable energy-harvesting computer systems is far-reaching," says Lucia. "Small satellites are proliferating and the space industry itself is expanding. If we can guarantee that even tiny, energy-harvesting satellites operate without interruption, we can make it easier to conduct other scientific research in space. Further out, we may even see future applications like extraterrestrial natural resource discovery relying on this technology." More information: Learn more about Professor Lucia's research group here. Read the researcher's paper "Chain: Tasks and Channels for Reliable Intermittent Programs" here.