Lalit Manam

Lalit Manam

Postdoctoral Researcher Mitsubishi Electric Research Labs

Curriculum Vitae

Download my CV - Curriculum Vitae

Research Interests

I am currently interested in 3D reconstruction problems in computer vision, primarily on large-scale structure-from-motion and SLAM pipelines. My recent work has been on developing methods for efficient and reliable estimation of camera motions in structure-from-motion. Previously, I was involved in the development of impulse noise removal techniques from colour images.

Current Research

I am working on 3D reconstruction problems in computer vision, i.e. to build a 3D representation of a scene from images. Depending on the application, it can be broadly classified as follows:

Depth cannot be recovered with a single image. Multiple images of the same scene are required to capture its 3D structure. The geometry of both problems is the same, as both applications utilize cameras. Building a 3D representation requires estimating the orientation/viewing direction as well as the position of the cameras. My current work focuses on estimating orientations (also called rotation averaging) and positions of the cameras (called translation averaging).

Applications of SfM and SLAM algorithms also include 3D exploration of photo collections and phone-based augmented reality.

Experience

Education

Talks/Tutorials/Lectures

Teaching Assistantships