| Presentation | 2010-03-16 Improvement of Accuracy in Bayesian Hidden Markov Model Approach for Sports Event Detection Tomohiro YAZAKI, Toshie MISU, Yohei NAKADA, Shigeru MOTOI, Go KOBAYASHI, Takashi MATSUMOTO, Nobuyuki YAGI, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | The problem of detecting the occurrence of target events in a given data sequence can be found in many fields , such as signal processing, pattern recognition, and image processing. We propose an event prediction method by using the Bayesian hidden Markov Model (HMM) and apply it to a video dataset of an actual soccer game. In this method, time-series data on the players' position are estimated from the video dataset and modeled with the HMM. The accuracy of this event prediction method can be improved by modifying the duration-modeling capability of the HMM, since a geometric distribution is used as its duration probability. In this paper, we use a generalized HMM in the event prediction algorithm for improving the duration-modeling capability. In addition, we test the efficiency of the proposed method with a video dataset of J-league soccer matches. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Event detection / Sports video analysis / Generalized hidden Markov Model / Bayesian learning / Markov chain Monte Carlo |
| Paper # | PRMU2009-301,HIP2009-186 |
| Date of Issue |
| Conference Information | |
| Committee | HIP |
|---|---|
| Conference Date | 2010/3/8(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | |
| Vice Chair | |
| Secretary | |
| Assistant | |
| Paper Information | |
| Registration To | Human Information Processing (HIP) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Improvement of Accuracy in Bayesian Hidden Markov Model Approach for Sports Event Detection |
| Sub Title (in English) | |
| Keyword(1) | Event detection |
| Keyword(2) | Sports video analysis |
| Keyword(3) | Generalized hidden Markov Model |
| Keyword(4) | Bayesian learning |
| Keyword(5) | Markov chain Monte Carlo |
| 1st Author's Name | Tomohiro YAZAKI |
| 1st Author's Affiliation | Faculty of Science and Engineering, Waseda University() |
| 2nd Author's Name | Toshie MISU |
| 2nd Author's Affiliation | Science & Technology Research Laboratories, Japan Broadcasting Corporation(NHK) |
| 3rd Author's Name | Yohei NAKADA |
| 3rd Author's Affiliation | Faculty of Science and Engineering, Waseda University |
| 4th Author's Name | Shigeru MOTOI |
| 4th Author's Affiliation | Faculty of Science and Engineering, Waseda University |
| 5th Author's Name | Go KOBAYASHI |
| 5th Author's Affiliation | Faculty of Science and Engineering, Waseda University |
| 6th Author's Name | Takashi MATSUMOTO |
| 6th Author's Affiliation | Faculty of Science and Engineering, Waseda University |
| 7th Author's Name | Nobuyuki YAGI |
| 7th Author's Affiliation | Science & Technology Research Laboratories, Japan Broadcasting Corporation(NHK) |
| Date | 2010-03-16 |
| Paper # | PRMU2009-301,HIP2009-186 |
| Volume (vol) | vol.109 |
| Number (no) | 471 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |