Group Member

Dhruv Batra
PhD. Student

Personal Homepage:  http://www.ece.cmu.edu/~dbatra
Email:

Spectral clustering and eigenvector-based methods have become increasingly popular in segmentation and recognition. Although the choice of the pairwise similarity metric (or affinities) greatly influences the quality of the results, this choice is typically specified outside the learning framework. In this work, we present an algorithm to learn class-specific similarity functions. Mapping our problem in a Conditional Random Fields (CRF) framework enables us to pose the task of learning affinities as parameter learning in undirected graphical models. There are two significant advances over previous work. First, we learn the affinity between a pair of data-points as a function of a pairwise feature and (in contrast with previous approaches) the classes to which these two data-points were mapped, allowing us to work with a richer class of affinities. Second, our formulation provides a principled probabilistic interpretation for learning all of the parameters that define these affinities. Using ground truth segmentations and labellings for training, we learn the parameters with the greatest discriminative power (in an MLE sense) on the training data. We demonstrate the power of this learning algorithm in the setting of joint segmentation and recognition of object classes. Specifically, even with very simple appearance features, the proposed method achieves state-of-the-art performance on standard datasets.

[ project page ]

Space-Time Shapelets for Action Recognition

Recent works in action recognition have begun to treat actions as space-time volumes. This allows actions to be converted into 3-D shapes, thus converting the problem into that of volumetric matching. However, the special nature of the temporal dimension and the lack of intuitive volumetric features makes the problem both challenging and interesting. In a data-driven and bottom-up approach, we propose a dictionary of mid-level features called Space-Time Shapelets. This dictionary tries to characterize the space of local space-time shapes, or equivalently local motion patterns formed by the actions. Representing an action as a bag of these space-time patterns allows us to reduce the combinatorial space of these volumes, become robust to partial occlusions and errors in extracting spatial support. The proposed method is computationally efficient and achieves competitive results on a standard dataset.

[ project page ]

An Alternative Formulation for the Five Point Relative Pose Problem

The “Five Point Relative Pose Problem” is to find all possible camera configurations between two calibrated views of a scene given five point-correspondences. We take a fresh look at this well-studied problem with an emphasis on the parametrization of Essential Matrices used by various methods over the years. Using one of these parametrizations, a novel algorithm is proposed, in which the solution to the problem is encoded in a system of nine quadratic equations in six variables, and is reached by formulating this as a constrained optimization problem. We compare our algorithm with an existing 5-point method, and show our formulation to be more robust in the presence of noise.

[ project page ]

This website is maintained by Devi Parikh
Copyright © 2001 Advanced Multimedia Processing Lab. All rights reserved.
Revised: March 09, 2008