Face recognition at a distance is a challenging and important law-enforcement surveillance problem, with low image resolution and blur contributing to the difficulties. We present a method for combining a sequence of video frames of a subject in order to create a restored image of the face with reduced blur. A generic Active AppearanceModel of face shape and appearance is used for registration....

This paper presents a method for obtaining accurate dense elevation and appearance models of terrain using a single camera on-board an aerial platform, which has many applications including geographical information systems, robot path planning, immersion and visualization, and surveying for scientific purposes such as watershed analysis. When given geo-registered images, the method can compute ...

David Kroetsch Master of Applied Science University of Waterloo September 2007 Multi-frame Measurement Fusion for State Estimation Simultaneous Localization and Mapping (SLAM) is a technique used by robots and autonomous vehicles to build up a map while operating in an unknown environment, while at the same time keeping track of a current position estimate. This process is not as straightforwar...

This paper proposes an approach, named phonetic context embedding, to model phonetic context effects for deep neural network hidden Markov model (DNN-HMM) phone recognition. Phonetic context embeddings can be regarded as continuous and distributed vector representations of context-dependent phonetic units (e.g., triphones). In this work they are computed using neural networks. First, all phone ...

Supplementing log filter-bank energies with i-vectors is a popular method for adaptive training of deep neural network acoustic models. While offline i-vectors (the target utterance or other relevant adaptation material is available for i-vector extraction prior to decoding) have been well studied, there is little analysis of online i-vectors and their robustness in multi-user scenarios where s...

We consider estimating a random vector from its measurements in a fusion frame, in presence of noise and subspace erasures. A fusion frame is a collection of subspaces, for which the sum of the projection operators onto the subspaces is bounded below and above by constant multiples of the identity operator. We first consider the linear minimum mean-squared error (LMMSE) estimation of the random...

Recovering structure from motion even using information from multiple image frames is diicult, in part because motion error can introduce large, correlated errors in the structure estimate. A method is proposed for recursively recovering structure from motion that can deal with this problem. Encouraging results on real images and synthetic data are presented.

The past few years have witnessed great success in applying deep learning to enhance the quality of compressed image/video. The existing approaches mainly focus on enhancing the quality of a single frame, ignoring the similarity between consecutive frames. In this paper, we investigate that heavy quality fluctuation exists across compressed video frames, and thus low quality frames can be enhan...

This paper develops a model for automated matching of vehicles between aerial photos. We exploit the overlap from frame-to-frame in a sequence of aerial photos. The problem, in essence, is to determine if vehicles A, B, C, . . . in one photo, match vehicles X, Y, and Z in a temporally lagged second photo. The method is specifically applied to large trucks (tractor–trailer combinations). If a ma...