IC Colloquium : A variety of manifold methods in Computer Vision
Event details
| Date | 22.09.2015 |
| Hour | 16:15 › 17:30 |
| Location | |
| Category | Conferences - Seminars |
By : Richard Hartley - Australian National University
Abstract :
I will talk about uses of Riemannian manifolds in computer vision, Examples of manifolds in CV include Kendall's Shape manifold, Grassman manifolds (used in recognition) the manifold of rotations, SO3, and positive-definite matrices. I will focus on two topics: kernel-learning methods on manifolds and optimization on manifolds.
The identification of kernels on manifolds allows us to apply techniques such as kernel-SVM and kernel dictionary-learning on spaces such as the Grassman manifold, with excellent results. Optimization methods directly on manifolds allow us to approach different geometric problems through techniques such as rotation averaging with different robust cost functions (Huber, L1)
Bio :
Richard Hartley is a member of the computer vision group in the Research School of Engineering, at the Australian National University, where he has been since January, 2001. He is a joint leader of the Computer Vision group in NICTA, a government funded research laboratory.
Dr. Hartley worked at the General Electric Research and Development Center from 1985 to 2001, working first in VLSI design, and later in computer vision. He became involved with Image Understanding and Scene Reconstruction working with GE's Simulation and Control Systems Division.
In 1991, he began an extended research effort in the area of applying projective geometry techniques to reconstruction using calibrated and semi-calibrated cameras. This research direction was one of the dominant themes in computer vision research throughout the 1990s. In 2000, he co-authored (with Andrew Zisserman) a book on Multiview Geometry in Computer Vision, summarizing the previous decade’s research in this area.
More information
Abstract :
I will talk about uses of Riemannian manifolds in computer vision, Examples of manifolds in CV include Kendall's Shape manifold, Grassman manifolds (used in recognition) the manifold of rotations, SO3, and positive-definite matrices. I will focus on two topics: kernel-learning methods on manifolds and optimization on manifolds.
The identification of kernels on manifolds allows us to apply techniques such as kernel-SVM and kernel dictionary-learning on spaces such as the Grassman manifold, with excellent results. Optimization methods directly on manifolds allow us to approach different geometric problems through techniques such as rotation averaging with different robust cost functions (Huber, L1)
Bio :
Richard Hartley is a member of the computer vision group in the Research School of Engineering, at the Australian National University, where he has been since January, 2001. He is a joint leader of the Computer Vision group in NICTA, a government funded research laboratory.
Dr. Hartley worked at the General Electric Research and Development Center from 1985 to 2001, working first in VLSI design, and later in computer vision. He became involved with Image Understanding and Scene Reconstruction working with GE's Simulation and Control Systems Division.
In 1991, he began an extended research effort in the area of applying projective geometry techniques to reconstruction using calibrated and semi-calibrated cameras. This research direction was one of the dominant themes in computer vision research throughout the 1990s. In 2000, he co-authored (with Andrew Zisserman) a book on Multiview Geometry in Computer Vision, summarizing the previous decade’s research in this area.
More information
Practical information
- General public
- Free
- This event is internal
Contact
- Host : Pascal Fua