Alan Brunton

Belief Propagation on the GPU for Stereo Vision

The power of Markov random field formulations of lowlevel vision problems, such as stereo, has been known for some time. However, recent advances, both algorithmic and in processing power, have made their application practical. This paper presents a novel implementation of Bayesian belief propagation for graphics processing units found in most modern desktop and notebook computers, and applies it to the stereo problem. The stereo problem is used for comparison to other BP algorithms.

Image-based navigation in real environments using panoramas

We present a system for virtual navigation in real environments using image-based panorama rendering. Multiple overlapping images are captured using a Point Grey Ladybug camera and a single cube-aligned panorama image is generated for each capture location. Panorama locations are connected in a graph topology and registered with a 2D map for navigation. A real-time image-based viewer renders individual 360-degree panoramas using graphics hardware acceleration. Real-world navigation is performed by traversing the graph and loading new panorama images. The system contains a user-friendly interface and supports standard input and display or a head-mounted display with an inertial tracking device.

Pushing the Limits of 3D Color Printing: Error Diffusion with Translucent Materials

Accurate color reproduction is important in many applications of 3D printing, from design prototypes to 3D color copies or portraits. Although full color is available via other technologies, multi-jet printers have greater potential for graphical 3D printing, in terms of reproducing complex appearance properties. However, to date these printers cannot produce full color, and doing so poses substantial technical challenges, from the shear amount of data to the translucency of the available color materials. In this article, we propose an error diffusion halftoning approach to achieve full color with multi-jet printers, which operates on multiple isosurfaces or layers within the object. We propose a novel traversal algorithm for voxel surfaces, which allows the transfer of existing error diffusion algorithms from 2D printing. The resulting prints faithfully reproduce colors, color gradients and fine-scale details.

Review of statistical shape spaces for 3D data with comparative analysis for human faces

Abstract With systems for acquiring 3D surface data being evermore commonplace, it has become important to reliably extract specific shapes from the acquired data. In the presence of noise and occlusions, this can be done through the use of statistical shape models, which are learned from databases of clean examples of the shape in question. In this paper, we review, analyze and compare different statistical models: from those that analyze the variation in geometry globally to those that analyze the variation in geometry locally. We first review how different types of models have been used in the literature, then proceed to define the models and analyze them theoretically, in terms of both their statistical and computational aspects. We then perform extensive experimental comparison on the task of model fitting, and give intuition about which type of model is better for a few applications. Due to the wide availability of databases of high-quality data, we use the human face as the specific shape we wish to extract from corrupted data.

Multilinear Wavelets: A Statistical Shape Space for Human Faces

We present a statistical model for 3D human faces in varying expression, which decomposes the surface of the face using a wavelet transform, and learns many localized, decorrelated multilinear models on the resulting coefficients. Using this model we are able to reconstruct faces from noisy and occluded 3D face scans, and facial motion sequences. Accurate reconstruction of face shape is important for applications such as tele-presence and gaming. The localized and multi-scale nature of our model allows for recovery of fine-scale detail while retaining robustness to severe noise and occlusion, and is computationally efficient and scalable. We validate these properties experimentally on challenging data in the form of static scans and motion sequences. We show that in comparison to a global multilinear model, our model better preserves fine detail and is computationally faster, while in comparison to a localized PCA model, our model better handles variation in expression, is faster, and allows us to fix identity parameters for a given subject.

Segmenting animated objects into near-rigid components

We present a novel approach to solve the problem of segmenting a sequence of animated objects into near-rigid components based on k given poses of the same non-rigid object. We model the segmentation problem as a clustering problem in dual space and find near-rigid segments with the property that segment boundaries are located at regions of large deformation. The presented approach is asymptotically faster than previous approaches that achieve the same property and does not require any user-specified parameters. However, if desired, the user may interactively change the number of segments. We demonstrate the practical value of our approach using experiments.

Filling holes in triangular meshes by curve unfolding

We propose a novel approach to automatically fill holes in triangulated models. Each hole is filled using a minimum energy surface that is obtained in three steps. First, we unfold the hole boundary onto a plane using energy minimization. Second, we triangulate the unfolded hole using a constrained Delaunay triangulation. Third, we embed the triangular mesh as a minimum energy surface in ℝ3. The running time of the method depends primarily on the size of the hole boundary and not on the size of the model, thereby making the method applicable to large models. Our experiments demonstrate the applicability of the algorithm to the problem of filling holes bounded by highly curved boundaries in large models.

Estimation of human body shape and posture under clothing

Estimating the body shape and posture of a dressed human subject in motion represented as a sequence of (possibly incomplete) 3D meshes is important for virtual change rooms and security. To solve this problem, statistical shape spaces encoding human body shape and posture variations are commonly used to constrain the search space for the shape estimate. In this work, we propose a novel method that uses a posture-invariant shape space to model body shape variation combined with a skeleton-based deformation to model posture variation. Our method can estimate the body shape and posture of both static scans and motion sequences of human body scans with clothing that fits relatively closely to the body. In case of motion sequences, our method takes advantage of motion cues to solve for a single body shape estimate along with a sequence of posture estimates. We apply our approach to both static scans and motion sequences and demonstrate that using our method, higher fitting accuracy is achieved than when using a variant of the popular SCAPE model [2,18] as statistical model.

Real-time video watermarking on programmable graphics hardware

In this paper, we propose a real-time video watermarking system on programmable graphics hardware. Real-time video watermarking is important to the use of digital video in legal proceedings, security surveillance, new reportage and commercial video transactions. The watermarking scheme implemented here is based on Wongs scheme for image watermarking, and is designed to detect and localize any change in the pixels of any frame of the incoming video stream. We implement this scheme for real-time operation on programmable graphics hardware. The graphics processing units (GPUs) found on many modern commodity-level graphics cards have the ability to execute application-defined sequences of instructions on not only geometric primitives, defined by vertices, but also on image or texture fragments mapped to rasterized geometric primitives. These fragment programs, also known as fragment or pixel shaders, execute in hardware and in parallel on the GPU for each fragment, or pixel, that is rendered, making the GPU well suited for image and video processing. We illustrate real-time performance, low perceptibility, and good bit-error rates and localization by way of a general testing framework that allows straightforward testing of any video watermarking system implemented on programmable graphics hardware

A low-dimensional representation for robust partial isometric correspondences computation

Intrinsic shape matching has become the standard approach for pose invariant correspondence estimation among deformable shapes. Most existing approaches assume global consistency. While global isometric matching is well understood, only a few heuristic solutions are known for partial matching. Partial matching is particularly important for robustness to topological noise, which is a common problem in real-world scanner data. We introduce a new approach to partial isometric matching based on the observation that isometries are fully determined by local information: a map of a single point and its tangent space fixes an isometry. We develop a new representation for partial isometric maps based on equivalence classes of correspondences between pairs of points and their tangent-spaces. We apply our approach to register partial point clouds and compare it to the state-of-the-art methods, where we obtain significant improvements over global methods for real-world data and stronger guarantees than previous partial matching algorithms.