Carl Olsson
Carl Olsson is a senior researcher in the Image Analysis and Computer Vision group lead by Prof. Fredrik Kahl. I his research he focuses on 3D reconstruction from images. Given a number of images of an object or scene the goal is get a computer to automatically create a 3D model of the scene as well as determine the camera placements when the images where captured. This technology is for example useful for autonomous vehicles that need to build maps to navigate their environment.
Showing 23 publications
Learning Structure-from-Motion with Graph Attention Networks
expOSE: Accurate Initialization-Free Projective Factorization using Exponential Regularization
AN UNBIASED APPROACH TO LOW RANK RECOVERY
Bias Versus Non-Convexity in Compressed Sensing
On the Tightness of Semidefinite Relaxations for Rotation Estimation
Radial Distortion Invariant Factorization for Structure from Motion
A quasiconvex formulation for radial cameras
Rotation Averaging with the Chordal Distance: Global Minimizers and Strong Duality
Bilinear Parameterization for Non-Separable Singular Value Penalties
Relaxations for Non-Separable Cardinality/Rank Penalties
Parameterization of Ambiguity in Monocular Depth Prediction
Differentiable Fixed-Rank Regularisation using Bilinear Parameterization
A Unified Optimization Framework for Low-Rank Inducing Penalties
An unbiased approach to compressed sensing
Accurate Optimization of Weighted Nuclear Norm for Non-Rigid Structure from Motion
Global Optimality for Point Set Registration Using Semidefinite Programming
Bilinear parameterization for differentiable rank-regularization
Combining Depth Fusion and Photometric Stereo for Fine-Detailed 3D Models
Rotation Averaging and Strong Duality
Compact Matrix Factorization with Dependent Subspaces
Non-convex Rank/Sparsity Regularization and Local Minima
Long-term 3D Localization and Pose from Semantic Labellings
A Non-Convex Relaxation for Fixed-Rank Approximation
Download publication list
You can download this list to your computer.
Filter and download publication list
As logged in user (Chalmers employee) you find more export functions in MyResearch.
You may also import these directly to Zotero or Mendeley by using a browser plugin. These are found herer:
Zotero Connector
Mendeley Web Importer
The service SwePub offers export of contents from Research in other formats, such as Harvard and Oxford in .RIS, BibTex and RefWorks format.
Showing 3 research projects
Learning and Leveraging Rich Priors for Factorization Problems
Optimization Methods with Performance Guarantees for Subspace Learning
Semantic Mapping and Visual Navigation for Smart Robots