- 著者
- Yasuhisa FUJII Kazumasa YAMAMOTO Seiichi NAKAGAWA
- 出版者
- The Institute of Electronics, Information and Communication Engineers
- 雑誌
- IEICE Transactions on Information and Systems (ISSN:09168532)
- 巻号頁・発行日
- vol.E95.D, no.4, pp.1101-1111, 2012-04-01 (Released:2012-04-01)
- 参考文献数
- 30
- 被引用文献数
- 2
This paper presents a novel method for improving the readability of automatic speech recognition (ASR) results for classroom lectures. Because speech in a classroom is spontaneous and contains many ill-formed utterances with various disfluencies, the ASR result should be edited to improve the readability before presenting it to users, by applying some operations such as removing disfluencies, determining sentence boundaries, inserting punctuation marks and repairing dropped words. Owing to the presence of many kinds of domain-dependent words and casual styles, even state-of-the-art recognizers can only achieve a 30-50% word error rate for speech in classroom lectures. Therefore, a method for improving the readability of ASR results is needed to make it robust to recognition errors. We can use multiple hypotheses instead of the single-best hypothesis as a method to achieve a robust response to recognition errors. However, if the multiple hypotheses are represented by a lattice (or a confusion network), it is difficult to utilize sentence-level knowledge, such as chunking and dependency parsing, which are imperative for determining the discourse structure and therefore imperative for improving readability. In this paper, we propose a novel algorithm that infers clean, readable transcripts from spontaneous multiple hypotheses represented by a confusion network while integrating sentence-level knowledge. Automatic and manual evaluations showed that using multiple hypotheses and sentence-level knowledge is effective to improve the readability of ASR results, while preserving the understandability.
- 著者
- Ikki SHIMIZU Keizo FURUYA Haruhiko OSAWA Yasuhisa FUJII Hideichi MAKINO
- 出版者
- The Japan Endocrine Society
- 雑誌
- Endocrine Journal (ISSN:09188959)
- 巻号頁・発行日
- vol.54, no.3, pp.477-480, 2007 (Released:2007-07-19)
- 参考文献数
- 13
- 被引用文献数
- 6 7
Insulin-induced localized lipoatrophy is a well-known localized side effect. Although an immune process has been suggested as its etiology, no definitive evidence has been reported to show that insulin is involved. Here, we report the first evidence for the phagocytosis of insulin by histiocytes as a very early stage of lipoatrophy, which was reproducible in two different lobular panniculitis tissues from a 71 year-old male patient. He had taken a subcutaneous insulin injection in his arms because of a sight disturbance. Since these subcutaneous tumors were likely due to inflammation by insulin, a biopsy sample was taken from the subcutaneous tumor of his right arm with his consent. The primary antibodies for insulin (1 : 200) and CD68 (1 : 50) were obtained from DAKO (guinea pig anti-insulin and mouse anti-human CD68). HE staining revealed the infiltration of mononuclear cells and histiocytes into the subcutaneous fat tissue, and some parts of this tissue had fibrosis with rich collagen fibers. These findings are consistent with a lobular panniculitis. Some histiocytes contained intracellular substances with a positive immunoreactivity to insulin. This activity was reduced when the anti-insulin antibody was preincubated with an excess amount of insulin antigen. The same substances were also positive to CD68. Thus, the phagocytosis of insulin by histiocytes appears to occur in this region. Therefore, the activation of subcutaneous macrophages by the complex of insulin and insulin antibodies may account for the initial autoimmune process.