A tutorial facial landmarks detection. I discussed the AAMs facial landmarks detection and implemented it. With this knowledge, We read a few good papers.
Category: applications
Forward and Inverse Kinematics
A simple introduction to the robot arm problem.
How to Land a Rocket? Part 3, Exploit the Structure
Improve the running time of the rocket landing problem by converting the KKT conditions into a block-diagonal system and solve it in O(n) time.
Variable Elimination, Smoothing, and Marginalization Explained
1. Explain the variable elimination in an oversimplified way. 2. Explain Fixed-lag smoothing by variable elimination. 3. Discuss a IROS2018 best paper candidate. 4. Some discussion of ISAM2, Semidefinite Programming.
How to Land a rocket? Part 2, Constrained Optimization
Apply constraints to the rocket landing problem. This post gives a introduction to Primal-Dual Interior Point methods.
How to Land a Rocket? Part 1, Problem Definition
Land a rocket 1. Problem formulation and simple tricks.
Lucas-Kanade on Manifolds
Lucas-kanade with SE(2) motion model. We will derive the equations and implement the tracker in C++.
The Covariance of Optimization Problems
1. Show the duality between optimization and probability. 2. Derive the covariance for optimization problems. 3. Implement the covariance for GPS and Lucas-Kanade tracker.
Time Calibration
We will 1. discuss the time-calibration problem. 2. Time calibration's connection to Lucas-Kanade. 3. IROS 2018 best paper. Apply to visual odometry
Lucas–Kanade Tracker
Derive Lucas-Kanade as a least squares problem. Implementation in C++.