A wearable and high-precision sensor for sound signal acquisition and recognition was fabricated from thin films of specially designed graphene woven fabrics （GWFs）. Upon being stretched, a high density of random cr...A wearable and high-precision sensor for sound signal acquisition and recognition was fabricated from thin films of specially designed graphene woven fabrics （GWFs）. Upon being stretched, a high density of random cracks appears in the network, which decreases the current pathways, thereby increasing the resistance. Therefore, the film could act as a strain sensor on the human throat in order to measure one＇s speech through muscle movement, regardless of whether or not a sound is produced. The ultra-high sensitivity allows for the realization of rapid and low-frequency speech sampling by extracting the signature characteristics of sound waves. In this study, representative signals of 26 English letters, typical Chinese characters and tones, and even phrases and sentences were tested, revealing obvious and characteristic changes in resistance. Furthermore, resistance changes of the graphene sensor responded perfectly with pre-recorded sounds. By combining artificial intelligence with digital signal processing, we expect that, in the future, this graphene sensor will be able to successfully negotiate complex acoustic systems and large quantities of audio data.展开更多
Statistical machine translation for low-resource language suffers from the lack of abundant training corpora. Several methods, such as the use of a pivot language, have been proposed as a bridge to translate from one ...Statistical machine translation for low-resource language suffers from the lack of abundant training corpora. Several methods, such as the use of a pivot language, have been proposed as a bridge to translate from one language to another. However, errors will accumulate during the extensive translation pipelines. In this paper, we propose an approach to low-resource language translation by exploiting the pronunciation correlations between languages. We find that the pronunciation features can improve both Chinese-Vietnamese and Vietnamese- Chinese translation qualities. Experimental results show that our proposed model yields effective improvements, and the translation performance （bilingual evaluation understudy score） is improved by a maximum value of 1.03.展开更多
基金This work was supported by Beijing Science and Technology Program （No. D141100000514001）, National Natural Science Foundation of China （No. 51372133）, and National Program on Key Basic Research Project （Nos. 2011CB013000 and 2014CB932401）
文摘A wearable and high-precision sensor for sound signal acquisition and recognition was fabricated from thin films of specially designed graphene woven fabrics （GWFs）. Upon being stretched, a high density of random cracks appears in the network, which decreases the current pathways, thereby increasing the resistance. Therefore, the film could act as a strain sensor on the human throat in order to measure one＇s speech through muscle movement, regardless of whether or not a sound is produced. The ultra-high sensitivity allows for the realization of rapid and low-frequency speech sampling by extracting the signature characteristics of sound waves. In this study, representative signals of 26 English letters, typical Chinese characters and tones, and even phrases and sentences were tested, revealing obvious and characteristic changes in resistance. Furthermore, resistance changes of the graphene sensor responded perfectly with pre-recorded sounds. By combining artificial intelligence with digital signal processing, we expect that, in the future, this graphene sensor will be able to successfully negotiate complex acoustic systems and large quantities of audio data.
基金supported by the National key Basic Research and Development(973)Program of China(No.2013CB329303)the National Natural Science Foundation of China(Nos.61502035,61132009,and 61671064)Beijing Advanced Innovation Center for Imaging Technology(No.BAICIT-2016007)
文摘Statistical machine translation for low-resource language suffers from the lack of abundant training corpora. Several methods, such as the use of a pivot language, have been proposed as a bridge to translate from one language to another. However, errors will accumulate during the extensive translation pipelines. In this paper, we propose an approach to low-resource language translation by exploiting the pronunciation correlations between languages. We find that the pronunciation features can improve both Chinese-Vietnamese and Vietnamese- Chinese translation qualities. Experimental results show that our proposed model yields effective improvements, and the translation performance （bilingual evaluation understudy score） is improved by a maximum value of 1.03.
