Publications-Proceedings

Article View/Open

Publication Export

Google ScholarTM

NCCU Library

Citation Infomation

Related Publications in TAIR

題名 Recognizing live fish species by hierarchical partial classification based on the exponential benefit
作者 沈錳坤
Chuang, Meng-Che
Hwang, Jenq-Neng
Kuo, Fang-Fei
Shan, Man-Kwan
Williams, Kresimir
貢獻者 資訊科學系
關鍵詞 exponential benefit; feature extraction; hierarchical partial classification; live fish recognition; underwater imagery
日期 2014-01
上傳時間 16-Aug-2017 17:04:51 (UTC+8)
摘要 Live fish recognition in open aquatic habitats suffers from the high uncertainty in many of the data. To alleviate this problem without discarding those data, the system should learn a species hierarchy so that high-level labels can be assigned to ambiguous data. In this paper, a systematic hierarchical partial classification algorithm is therefore proposed for underwater fish species recognition. Partial classification is applied at each level of the species hierarchy so that the coarse-to-fine categorization stops once the decision confidence is low. By defining the exponential benefit function, we formulate the selection of decision threshold as an optimization problem. Also, attributes from important fish anatomical parts are focused to generate discriminative feature descriptors. Experiments show that the proposed method achieves an accuracy up to 94%, with partial decision rate less than 5%, on underwater fish images with high uncertainty and class imbalance. © 2014 IEEE.
關聯 2014 IEEE International Conference on Image Processing, ICIP 2014,5232-5236
資料類型 conference
DOI http://dx.doi.org/10.1109/ICIP.2014.7026059
dc.contributor 資訊科學系zh_Tw
dc.creator (作者) 沈錳坤zh_TW
dc.creator (作者) Chuang, Meng-Cheen_US
dc.creator (作者) Hwang, Jenq-Nengen_US
dc.creator (作者) Kuo, Fang-Feien_US
dc.creator (作者) Shan, Man-Kwanen_US
dc.creator (作者) Williams, Kresimiren_US
dc.date (日期) 2014-01en_US
dc.date.accessioned 16-Aug-2017 17:04:51 (UTC+8)-
dc.date.available 16-Aug-2017 17:04:51 (UTC+8)-
dc.date.issued (上傳時間) 16-Aug-2017 17:04:51 (UTC+8)-
dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/111988-
dc.description.abstract (摘要) Live fish recognition in open aquatic habitats suffers from the high uncertainty in many of the data. To alleviate this problem without discarding those data, the system should learn a species hierarchy so that high-level labels can be assigned to ambiguous data. In this paper, a systematic hierarchical partial classification algorithm is therefore proposed for underwater fish species recognition. Partial classification is applied at each level of the species hierarchy so that the coarse-to-fine categorization stops once the decision confidence is low. By defining the exponential benefit function, we formulate the selection of decision threshold as an optimization problem. Also, attributes from important fish anatomical parts are focused to generate discriminative feature descriptors. Experiments show that the proposed method achieves an accuracy up to 94%, with partial decision rate less than 5%, on underwater fish images with high uncertainty and class imbalance. © 2014 IEEE.en_US
dc.format.extent 209 bytes-
dc.format.mimetype text/html-
dc.relation (關聯) 2014 IEEE International Conference on Image Processing, ICIP 2014,5232-5236en_US
dc.subject (關鍵詞) exponential benefit; feature extraction; hierarchical partial classification; live fish recognition; underwater imageryen_US
dc.title (題名) Recognizing live fish species by hierarchical partial classification based on the exponential benefiten_US
dc.type (資料類型) conference
dc.identifier.doi (DOI) 10.1109/ICIP.2014.7026059
dc.doi.uri (DOI) http://dx.doi.org/10.1109/ICIP.2014.7026059