International Workshop on Particle Filters for Data Assimilation
Co-organizers: Research and Development Center for Data Assimilation, ISM
Research Meeting on Particle Filters
Sponsored by the ISM Cooperative Research Program (2012- ISM・CRP -2017)
February 7, 2013, 1pm
Seminar Room D313, the Institute of Statistical Mathematics
（10-3, Midori-cho, Tachikawa, Tokyo 190-8562, JAPAN）
http://www.ism.ac.jp/access/index_e.html
1. Special Lecture
"Implicit Sampling: Theory and Implementation"
Brad Weir, Oregon State University (Oregon State University, http://people.oregonstate.edu/~weirb/)
This talk presents an algorithm that simultaneously estimates the state evolution and parameters of a non-linear stochastic model based on noisy observations of the state. The method approximates the probability distribution of the model solutions and parameters with an ensemble of sample values called particles: for each individual particle, it first determines the most probable values given the observed data, then samples around those values. This approach has a strong theoretical foundation, applies to non-Gaussian distributions, and can be implemented as a smoother or filter to estimate any number of model parameters, initial conditions, and/or model error covariances. It is referred to as "implicit" because it updates the particles without forming a predictive distribution of forward model integrations. The focus of the talk is the theoretical basis for the method and the practical issues involved in its implementation when computational time and storage are limited.
2. Workshop I (in English)
"Two ways of representation of the probability density function for ensemble-based data assimilation"
Shin'ya Nakano, ISM
"Influenza epidemic simulation in an urban area"
Masaya M. Saito, ISM
"Time-series modeling of tide gauge records using the particle filter for monitoring crustal activities related to oceanic trench earthquakes near Japan"
Hiromichi Nagao, ISM
3. Workshop II (in Japanese)
「ピクセル比率を尤度とするパーティクルフィルタ動画像追跡のGPGPUへの並列実装」
"Parallel implementation of visual tracking by particle filter with pixel ratio likelihood to GPGPU"
生駒哲一（九州工業大学）
Norikazu Ikoma (Kyushu Institute of Technology, Japan)
「疎に配置された複数カメラ動画像からの広域シーン中の移動対象追跡」
"On a moving target tracking in board scene from video images of sparsely located multiple cameras"
宮田太郎（九州工業大学大学院）
Taro Miyata (Kyushu Institute of Technology, Japan)
「交差後予測に基づくパーティクルフィルタを用いた外観類似物体同士の交差に頑健な物体追跡」
"Robust Object Tracking for Intersections of Objects with Similar Pattern using Prediction After Intersection based Particle Filter"
○椎名雄飛、池永剛（早稲田大学）
*Yuhi Shiina, Takeshi Ikenaga (Waseda University, Japan)
「反復画像再構成法による複数海洋レーダ画像合成」
"Iterative image reconstruction method for multiple radar image systems"
○小橋昌司，三田北斗，中川和也，西山浩二，前野 仁，倉本 圭，畑 豊（兵庫県立大学）
Hokuto Mita, *Syoji Kobashi, Kazuya Nakagawa, Kohji Nishiyama, Hitoshi Maeno, Kei Kuramoto, Yutaka Hata (Univ. of Hyogo, Japan)
「一般状態空間モデル解析による方向波スペクトルの逆推定」
"Inverse estimation for directional wave spectra based on the general state space modeling procedure"
寺田大介（水産総合研究センター 水産工学研究所）
Daisuke Terada (National Research Institute of Fisheries Engineering, Japan)
「確率的セルオートマトン型群集モデルに対するデータ同化」
"Data assimilation of probabilistic cell automaton for crowd modeling"
田代祐志，○川本一彦（千葉大学）
Yuji Tashiro, *Kazuhiko Kawamoto (Chiba University, Japan)
＝＝＝
