M Codes
I am working on coding algorithms related to the particle filter (PF)/smoothers and relevants for target tracking (including random finite set (RFS)-based multi-target tracking), as well as on data-driven/model-free methods for target tracking. Below are some of my achievements on which you may have better implementation.
-------------------------------------------------
A batch of resampling methods for the PF that have been realized in Matlab as attached in separate .m files
Ref: T. Li, M. Bolic, P. Djuric, Resampling methods for particle filtering, IEEE Signal Processing Magazine, IEEE Signal Processing Magazine, 2015, vol.32, no.3, pp. 70-86. @ IEEE Xplore
---------------------------------------------------
2. The Matlab codes of a series of Particle filters and smoothers as well as some Particle methods for parameter estimation.
---------------------------------------------------
3. I am currently working on target tracking and sensor fusion, with emphsis on RFS-based multi-target tracking, such as PF/SMC-PHD filter and related.
----------------------------------------------------------
4. My favorite yet a bit crazy work in target tracking is to get rid of any filters but resort to data mining and learning algorithm for tracking! Please refer to the subpage and codes at C4F, FoT4STF,T-Constrained tracking, and O2 --> C4F --> F4S --> FoT4STF!
5. More recently, I am working the multi-sensor multitarget tracking based on the arithmatic average density fusion, via averaging the probability density in the single target case like Kalman filter and Student's t filter, or averaging the probability hypothesis density (PHD) of various RFS/LRFS filters. Codes for AA fusion are available at AA fusion and P/P Consensus- AA Fusion
Let the data speak!
There are some very nice codes for group/extended target tracking shared by Dr. Lyudmila Mihaylova @ matlab central and for change-detection models/ parameter estimation based EM/SMC shared by Prof. Arnaud Doucet. In addition is an excellent software, Lawrence Murray 's LibBi (avaiable @ http://www.indii.org/research) for state-space modelling and Bayesian inference on high-performance computer hardware including multi-core CPUs/GPUs and distributed-memory clusters. These methods include EKF, particle Markov chain Monte Carlo (PMCMC), SMC2 and some other optimization routines.
In addition, Prof. Vo has shared his Matlab Codes for RFS filters for multi-target tracking including the very popular methods like PHD/CPHD filters and Labelled Multi-Bernoulli filter.