Semantic Event Detection in Baseball Videos Based on a Two-layer Hidden Markov Model
碩士 === 雲林科技大學 === 資訊工程研究所 === 98 === In this paper, we proposed an event detection method in baseball videos based on a two-layer HMM (hidden Markov model), using high-level audio/video features. For the video part, we use eight kinds of semantic scenes detected from baseball videos in our previous...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | en_US |
Published: |
2010
|
Online Access: | http://ndltd.ncl.edu.tw/handle/25709687614592630464 |
id |
ndltd-TW-098YUNT5392011 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-098YUNT53920112015-10-13T18:58:57Z http://ndltd.ncl.edu.tw/handle/25709687614592630464 Semantic Event Detection in Baseball Videos Based on a Two-layer Hidden Markov Model 基於兩層隱藏式馬可夫模型用於棒球的事件偵測 Jyun-Jhang Huang 黃俊彰 碩士 雲林科技大學 資訊工程研究所 98 In this paper, we proposed an event detection method in baseball videos based on a two-layer HMM (hidden Markov model), using high-level audio/video features. For the video part, we use eight kinds of semantic scenes detected from baseball videos in our previous work. For the audio part, we extract the audio shots from corresponding video scenes, and cut an audio shot into N one-second clips. Then, the MFCC and ZCR of a one-second clip are extracted and fed into the SVM for classifying it as “acclaim” and “silence”. Based on the classification results, the type of an audio shot can be determined in the post-classification. Next, a two-layer HMM modified from the original HMM is used to combine video and audio features to detect baseball video events. Finally, the experimental results show, the two-layer HMM has good event detection accuracy. Yin-Fu Huang 黃胤傅 2010 學位論文 ; thesis 29 en_US |
collection |
NDLTD |
language |
en_US |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 雲林科技大學 === 資訊工程研究所 === 98 === In this paper, we proposed an event detection method in baseball videos based on a two-layer HMM (hidden Markov model), using high-level audio/video features. For the video part, we use eight kinds of semantic scenes detected from baseball videos in our previous work. For the audio part, we extract the audio shots from corresponding video scenes, and cut an audio shot into N one-second clips. Then, the MFCC and ZCR of a one-second clip are extracted and fed into the SVM for classifying it as “acclaim” and “silence”. Based on the classification results, the type of an audio shot can be determined in the post-classification. Next, a two-layer HMM modified from the original HMM is used to combine video and audio features to detect baseball video events. Finally, the experimental results show, the two-layer HMM has good event detection accuracy.
|
author2 |
Yin-Fu Huang |
author_facet |
Yin-Fu Huang Jyun-Jhang Huang 黃俊彰 |
author |
Jyun-Jhang Huang 黃俊彰 |
spellingShingle |
Jyun-Jhang Huang 黃俊彰 Semantic Event Detection in Baseball Videos Based on a Two-layer Hidden Markov Model |
author_sort |
Jyun-Jhang Huang |
title |
Semantic Event Detection in Baseball Videos Based on a Two-layer Hidden Markov Model |
title_short |
Semantic Event Detection in Baseball Videos Based on a Two-layer Hidden Markov Model |
title_full |
Semantic Event Detection in Baseball Videos Based on a Two-layer Hidden Markov Model |
title_fullStr |
Semantic Event Detection in Baseball Videos Based on a Two-layer Hidden Markov Model |
title_full_unstemmed |
Semantic Event Detection in Baseball Videos Based on a Two-layer Hidden Markov Model |
title_sort |
semantic event detection in baseball videos based on a two-layer hidden markov model |
publishDate |
2010 |
url |
http://ndltd.ncl.edu.tw/handle/25709687614592630464 |
work_keys_str_mv |
AT jyunjhanghuang semanticeventdetectioninbaseballvideosbasedonatwolayerhiddenmarkovmodel AT huángjùnzhāng semanticeventdetectioninbaseballvideosbasedonatwolayerhiddenmarkovmodel AT jyunjhanghuang jīyúliǎngcéngyǐncángshìmǎkěfūmóxíngyòngyúbàngqiúdeshìjiànzhēncè AT huángjùnzhāng jīyúliǎngcéngyǐncángshìmǎkěfūmóxíngyòngyúbàngqiúdeshìjiànzhēncè |
_version_ |
1718039629649674240 |