Select a calendar:
Filter May Events by Event Type:
SUNMONTUEWEDTHUFRISAT
University Calendar
Events for May
-
PhD Defense - Franziska Meier
Tue, May 03, 2016 @ 10:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
University Calendar
Title: Towards a Probabilistic Motor Skill Learning Framework
Location: RTH 406
Time: 10am - noon, May 3rd, 2016
PhD Candidate: Franziska Meier
Committee members:
Prof. Stefan Schaal (Chair)
Prof. Gaurav Sukhatme
Prof. Yan Liu
Prof. James Finley
Abstract:
While learning in robotics has always been seen as one of the
hallmarks to accomplish autonomous behaviors, so far, there is no coherent and robust approach to robot learning. For instance, realizing complex behaviors, such as manipulation skills, often involves a mixing and matching of planning, control, and learning modules, dominated by the insights of the robotics researcher, but not by a coherent design and/or algorithmic principle. Thus, most robot learning approaches have largely remained a proof-of-concept rather then a general research approach towards robot learning.
In this thesis we aim to move towards a motor skill learning framework coherently routed in probabilistic representations. The use of probabilistc graphical models for different learning modules can foster a principled combination of these modules to form an integrated approach to skill representations. Towards this goal I will present
contributions from two directions: Creating computationally efficient approximations of probabilistic graphical models and developing probabilistic solutions to problems in motor skill learning.
In the first part of my talk I will present our work on scaling up learning in graphical models such that the use of a complex graphical model -- as would be required for complex motor skill representations with perceptual coupling -- is feasible.
In the second part of the talk, I will then present probabilistic approaches to two subproblems of motor skill learning. First I will introduce a probabilistic version of dynamic movement primitives. With the help of this formulation we can implement online movement recognition and perform segmentation of complex skill sequences into movement primitives.
Finally, I will tackle the problem of learning internal models, exemplified by inverse dynamics learning. Having a good inverse dynamics model ensures that we can execute trajectories in an accurate yet compliant manner. I will present a real-time capable drifting Gaussian process approach to learning a local approximation of the
inverse dynamics model on the fly.
Location: Ronald Tutor Hall of Engineering (RTH) - 406
Audiences: Everyone Is Invited
Contact: Lizsl De Leon
-
PhD Defense - Guan Pang
Thu, May 05, 2016 @ 12:00 PM - 02:00 PM
Thomas Lord Department of Computer Science
University Calendar
Title: 3D Object Detection in Industrial Site Point Clouds
Location: SAL 322
Time: 12:00pm - 2:00pm, May 5th, 2016
PhD Candidate: Guan Pang
Committee members:
Prof. Ulrich Neumann (Chair)
Prof. Aiichiro Nakano
Prof. C.-C. Jay Kuo (Outside Member)
Abstract:
Detection of three dimensional (3D) objects in point clouds is a challenging problem. Existing methods either focus on a specific type of object or scene, or require prior segmentation, both of which are usually inapplicable on real-world industrial applications.
This thesis describe three methods to tackle the problem, with gradually improving performance and efficiency. The first is a general purpose 3D object detection method that combines Adaboost with 3D local features, without requirement for prior object segmentation. Experiments demonstrated competitive accuracy and robustness to occlusion, but this method suffers from limited rotation invariance. As an improvement, another method is presented with a multi-view detection approach that projects the 3D point clouds into several 2D depth images from multiple viewpoints, transforming the 3D problem into a series of 2D problems, which reduces complexity, stabilizes performance, and achieves rotation invariance. The problem is the huge amount of projected views and rotations that need to be individually detected, limiting the complexity and performance of 2D algorithm choice. Thus the third method is proposed to solve this with the introduction of convolutional neural network, because it can handle all viewpoints and rotations for the same class of object together, as well as predicting multiple classes of objects with the same network, without the need for individual detector for each object class. The detection efficiency is further improved by concatenating two extra levels of early rejection networks with binary outputs before the multi-class detection network.
3D object detection in point clouds is crucial for 3D industrial point cloud modeling. Prior efforts focus on primitive geometry, street structures or indoor objects, but industrial data has rarely been pursued. We integrate several algorithm components into an automatic 3D modeling system for industrial site point clouds, including modules for pipe modeling, plane classification and object detection, and solves the technology gaps revealed during the integration. The integrated system is able to produce classified models of large and complex industrial scenes with a quality that outperforms leading commercial software and comparable to professional hand-made models.
This thesis also describes an earlier work in multi-modal image matching which inspires later research in 3D object detection by 2D projections. Most existing 2D descriptors only work well on images of a single modality with similar texture. This proposal presents a novel basic descriptor unit called a Gixel, which uses an additive scoring method to sample surrounding edge information. Several Gixels in a circular array create the Gixel Array Descriptor, excelling in multi-modal image matching with dominant line features.
Location: Henry Salvatori Computer Science Center (SAL) - 322
Audiences: Everyone Is Invited
Contact: Lizsl De Leon
-
PhD Defense - Shay Deutsch
Fri, May 06, 2016 @ 02:30 PM - 04:30 PM
Thomas Lord Department of Computer Science
University Calendar
Title: Learning the Geometric Structure of High Dimensional Data using the Tensor Voting Graph
Location: SAL 322
Time: 2:30pm - 4:30pm, May 6th, 2016
PhD Candidate: Shay Deutsch
Committee members:
Prof. Gerard Medioni (Chair)
Prof. Aiichiro Nakano
Prof. Antonio Ortega (Outside Member)
Abstract:
This study addresses a range of fundamental problems in unsupervised manifold learning. Given a set of noisy points in a high dimensional space that lie near one or more possibly intersecting smooth manifolds, different challenges include learning the local geometric structure at each point, geodesic distance estimation, and clustering. These problems are ubiquitous in unsupervised manifold learning, and many applications in computer vision as well as other scientific applications would benefit from a principled approach to these problems.
In the first part of this thesis we present a hybrid local-global method that leverages the algorithmic capabilities of the Tensor Voting framework. However, unlike Tensor Voting, which can learn complex structures reliably only locally, our method is capable of reliably inferring the global structure of complex manifolds using a unique graph construction called the Tensor Voting Graph (TVG). This graph provides an efficient tool to perform the desired global manifold learning tasks such as geodesic distance estimation and clustering on complex manifolds, thus overcoming one of one of the main limitations of Tensor Voting as a strictly local approach. Moreover, we propose to explicitly and directly resolve the ambiguities near the intersections with a novel algorithm, which uses the TVG and the positions of the points near the manifold intersections.
In the second part of this thesis we propose a new framework for manifold denoising based processing in the graph Fourier frequency domain, derived from the spectral decomposition of the discrete graph Laplacian. The suggested approach, called MFD, uses the Spectral Graph Wavelet transform in order to perform non-iterative denoising directly in the graph frequency domain. To the best of our knowledge, MFD is the first attempt to use graph signal processing tools for manifold denoising on unstructured domains. We provide theoretical justification for our Manifold Frequency Denoising approach on unstructured graphs and demonstrate that for smooth manifolds the coordinate signals also exhibit smoothness. This is first demonstrated in the case of noiseless observations, by proving that manifolds with smoother characteristics creates more energy in the lower frequencies. Moreover, it is shown that higher frequency wavelet coefficients decay in a way that depends on the smoothness properties of the manifold, which is explicitly tied to the curvature properties. We then provide an analysis for the case of noisy points and a noisy graph, establishing results which tie the noisy graph Laplacian to the noiseless graph Laplacian characteristics, induced by the smoothness manifold properties and the graph construction properties.
Finally, the last part of this research merges the Manifold Frequency Denoising and the Tensor Voting Graph methods into a uniform framework, which allows us to denoise and analyze a general class of noisy manifolds with singularities also in the presence of outliers. We demonstrate that the limitation of the Spectral Graph Wavelets in its flexibility to analyze certain classes of graph signals can be overcome for manifolds with singularities using certain graph construction and regularization methods. This allows us to take into account global smoothness characteristics without over-smoothing in the manifold discontinuations (which correspond to high frequency bands of the Spectral Graph Wavelets), and moreover is robust to a large amount of outliers.
Location: Henry Salvatori Computer Science Center (SAL) - 322
Audiences: Everyone Is Invited
Contact: Lizsl De Leon
