ECura Project at the 3rd International Symposium of the ICTM Study Group on Audiovisual Ethnomusicology

 

ECura project local senior advisor Xin Weibo at the conference (photo by Su Xiaoyin)
ECura project local senior advisor Su Xiaoyin and PI Dr Qian Lijuan at the conference (photo by Xin Weibo)

ECura promotes the establishment of Yi Spoken language database

To prepare for the establishment of the Yi speech database, ECura’s postdoc researcher Dr. Jin Dai and project leader Dr. Lijuan Qian conducted fieldwork in China in June and July, respectively. During their stay, they visited Mayou village, Kunming city, and Shanghai. They engaged with scholars and industry experts on data collection, corpus translation, and technical architecture required for building the Yi spoken language database.

为筹建彝族语音数据库,ECura的博士后研究员戴进博士和项目主管钱丽娟博士分别于六月和七月在中国进行了田野考察。为准备筹建彝族语音数据库,两位项目组成员先后实地走访了马游村、昆明市和上海市,分别从建立数据库所需的数据收集,语料翻译,以及技术架构三个方面和相关的学者及业界专家进行交流。

Jin Dai established preliminary connections with Yi folk musician, technical expert and linguist. During the fieldwork in Mayou Village, Jin collected Meige’s audios from collaborative Yi community leader and folk musician Xiaowei Guo and discussed the thematic content, tune styles, and classification methods of these audios. These audios would serve as raw material for creating the Yi speech database. Later, in Kunming, Jin visited scholars from local universities, including Associate Professor Dr. Xingfu Li, an anthropological scholar and a Yi linguistic expert at Minzu University of Yunnan, and Associate Professor in computer sciences Dr. Zhiwen Tang from the University of Yunnan. During his conversation with Xingfu Li, Jin learned about dialectal variations, lexical categories, the notation of the International Phonetic Alphabet (IPA), and grammatical principles of the Yi language. This knowledge preparation was important for the project to understand the basic linguistic framework of Yi and be aware of the differences in language contexts between Yi and Mandarin. Subsequently, in her meeting with Zhiwen Tang, Jin received many suggestions on the basic steps needed to establish a speech recognition corpus, including details about hardware equipment, feasible plans, and expected timelines.


Jin Dai collected audios from Meige Studio, which serves as a practical training base for numerous schools (photo by Jin Dai)

戴进与学术界的技术专家和语言学家建立了初步联系。在马游村的田野考察期间,戴进从彝族社区负责人郭晓炜那里收集了梅葛音频录音,并讨论了这些录音的主题内容、曲调风格和分类方法。这些音频将作为创建彝族语音数据库的原始材料。随后,在云南,戴进还拜访了当地大学的学者,包括云南民族大学的民族学者和彝族语言专家李兴福副教授,以及云南大学计算机科学领域的汤智文副教授。在与李兴福的交谈中,戴进了解了彝族语言的方言变体、词汇类别、国际音标的标注以及语法原则。这些知识准备对于项目了解彝族语言的基本语言框架以及彝语和汉语普通话之间的语言背景差异非常重要。随后,在与汤智文的见面中,戴进得到了关于建立语音识别语料库的基本步骤的许多建议,包括硬件设备、可行计划和预期时间线等详细信息。

Lijuan Qian, Xingfu Li, and Xiaowei Guo discussed the construction of the Yi language database together (photo by Lijuan Qian)

Xiaoying Su showed the recorded videos of Yi’s traditional songs (photo by Lijuan Qian)

In early July, Lijuan Qian continued to follow up on Jin’s fieldwork. In Mayou, Lijuan continued the audio collection work initiated by Jin and collaborated with Xingfu Li and Xiaowei Guo, who demonstrated to Lijuan the specific methods of segmenting and splitting speech using software. Subsequently, Xiaoying Su from Yunnan Yingda Imaging Company, who has extensive experience in ethnic videography and management, showcased the operation of their Yuansheng studio’s WeChat account to Lijuan. After preparing for audio collection, Lijuan’s efforts extended to Shanghai. She met with Shanghai Yizhe Technology, a database technology service provider, to understand the frontier technologies required for establishing the Yi speech database. The two parties discussed how to construct a multi-modal, multilingual parallel corpus based on the Meige theme. This database aims to use Meige audios as a reference, documenting and organizing corresponding Chinese, English, International Phonetic Alphabet (IPA), and Yi script translations, along with annotations, images, and videos. Additionally, the database aims to develop a product for automatic Yi speech recognition and translation based on language model training technology in its later stages. This will enable audiences from different cultural backgrounds to understand the orally performed Meige, increasing the possibility of cultural exchange.

七月初,钱丽娟在马游村和昆明市继续跟进戴进的田野工作。在马游,钱丽娟继续跟进梅葛音频的收集工作,并与李兴福和郭晓炜见面,由他们演示按语素和单句拆分音频的步骤。之后,有丰富的民族影像学的拍摄和管理经验的云南映达影视文化传媒有限公司的粟小殷向ECura展示了公司主理的‘源生坊’视频公众号的运营状况。在做好音频收集的准备工作后,为了了解建立彝族语音数据库所需的前沿技术,钱丽娟与数据库技术服务商上海一者科技见面,双方讨论如何组建一个基于‘梅葛’主题的多模态多语平行语料库。该数据库以现今仍然存活和传唱的梅葛音频资料为参照,记录并整理出与彝语语素音频相对应的中文、英文、国际音标以及彝文文字,并添加注释,图片、视频等多模态内容。此外,该数据库在后期建设中拟开发一个基于语言模型训练技术的彝语自动语音识别和自动翻译的产品,使口头演唱的梅葛音乐能被来自不同文化背景的观众理解,增加文化交流的可能性。

Written and edited by Jin Dai

ECura Initiates the Project’s Multilingual Parallel Corpus

In July 2024, Dr Lijuan Qian visited Shanghai Yizhe Information Technology Co., Ltd., a provider of multilingual corpus and intelligent translation platform. The meeting was hosted by Mr. Jing Zhang, the CEO of Yizhe Tech, and attended by other senior executives. Dr. Qian and Yizhe Tech discussed the current status of the preservation of low-resourced languages, such as the Yi language. Additionally, they exchanged views on frontier AI-based technologies such as natural language processing and the use of automatic speech recognition of minority languages in the preservation of orally transmitted cultural heritage. This meeting not only laid the foundation for future cooperation between ECura and the industry but also opened new avenues for interdisciplinary academic exchanges in the digitization and preservation of low-resource languages and intangible cultural heritage.

Lijuan Qian meets with the CEO of Shanghai Yizhe Tech Mr. Jing Zhang (Photo by Lijuan Qian)
Lijuan Qian has a meeting with Yizhe Tech (photo by Lijuan Qian)

ECura visits Meige Studio Singing Traditional Meige Children’s song

The project assistant, Yalun Yang, representing the ECura Team, visited Mayou village on October 4th, 2024, during the national holiday. As the English Meige Children’s Song teacher, she has taught children in Mayou village online for five months. This meeting aims to meet the children in person and build a more robust relationship with the local people. Community leaders Xiaowei Guo and Rongxiu Luo gathered students into the classroom in Meige Studio, and Yalun gave a class to the students.

2024 年 10 月 4 日,项目助理杨雅伦代表 ECura 团队在国庆节期间走访了马游村。在此前的五个月里,她一直在网上教孩子们唱英文版的梅葛童谣。此次见面旨在与孩子们面对面交流授课,以此与当地居民建立更稳固的关系。社区负责人郭晓炜和罗荣秀将学生们召集到梅葛坊的教室里,由杨雅伦给学生们上课。

During the class, Yalun reviewed the songs the children had learned, including ‘To Carve a Wooden Trough’, ‘Hide and Seek’, ‘Breaking Lumpy Earth’ and so on. Students sang duets in both Yi and English by separating the class into two groups, which helped review the English and Yi lyrics. Then, the children learned the new song, ‘Come and Dance with Your Feet’. Then, the students performed the Hulusheng and invited Yalun to dance together. The children liked the in-person class and enjoyed the gathering time during the holiday. Before some students returning boarding school, they had the chance to catch up with the English and Yi Meige Children’s Song learning in time.

课堂上,杨雅伦回顾了孩子们在过去五个月里所学的儿歌,包括《挖木啦》、《捉迷藏》、《敲土垡》等。学生们分成两组,用彝语和英语对唱,同时复习了英语和彝语歌词。随后,孩子们学习了新歌《一起来跳脚》。下课后,孩子们又表演了葫芦笙,并邀请杨雅伦一起跳舞。孩子们很喜欢线下课堂,也很享受假期的聚会时光。在部分学生返回寄宿学校之前,也有机会及时复习了英语和彝语梅葛童谣。

Yalun also reviewed several Meige lyrics with Xiaowei Guo and Rongxiu Luo. She will continue to teach the English Meige Children’s Song online weekly. Guo Xiaowei and Luo Rongxiu recognised the ECura team’s teaching of the English Meige Children’s Songs and expressed their willingness to deepen their cooperation.

杨雅伦还和郭晓炜、罗荣秀一起订正了几首梅葛歌曲的歌词。她将继续每周在线上进行梅葛儿歌的教学。郭晓炜和罗荣秀对 ECura 团队教孩子们梅葛童谣英文版的做法表示认可,并表达了深化合作的意愿。

Meige Children’s Song English class in person (photo by Yalun)
Meige Children’s Song English class in person ( video provided by Yalun)

Written and edited by Yalun Yang

Preparation for the Miao Spoken Language Database: Collaborate with Xuede Li

The ECura project members connected with Xuede Li, the Miao folk specialist of Northwest Yunnan Miao Dialect Area, a month before with the help of local Miao music scholar Yunhui Yang. A preliminary agreement between the ECura team and Mr. Li was reached online to record a voice database. On 5th Nov. 2024, Yalun Yang represented the ECura team and visited Zhaotong City. During the four-day stay, Yalun visited Yunhui Yang, Mr. Li and his family, including his wife, Rongfen Tian. Xuede and Rongfen signed a paper agreement with the ECura team to collaborate on recording for the Miao Spoken language database, and then Yalun helped them set up a bank account for receiving the money from the ECura team. Considering that it is inconvenient for an elderly person to work for long hours staring at a mobile phone screen, the ECura team equipped Mr. Li’s computer with a good quality microphone after consultation and helped him set up the microphone parameters and recording software. Yalun also helped Mr. Li type ancient Miao song lyrics into his computer and taught him how to select vocabulary so that Mr. Li could better understand the focus and process of database recording.

Yalun is setting the microphone for Xuede Li (Photo by Rongfen Tian) 
Yalun Yang is signing the contract with Xuede Li (left) and Rongfen Tian(Middle). (Photo by Yunhui Yang) 

Xuede Li invited Yalun and Yunhui to his hometown village, A Chu Lu. There, he introduced Yalun to the local musicians and participated in a local music performance.The ECura team’s collaboration with Mr Li Xuede shows the determination and action to build a spoken language database of ethnic minorities. In the future, we will also continue to give technical support to Mr Li and increase our recording and observation of the local Miao culture.

2024年11月5日,项目助理杨雅伦代表ECura团队来到昭通市开展了为期四天的工作。ECura团队于此前一个月通过苗族音乐学者杨云惠线上联系上了李学德老师,李学德是滇东北次方言区的民俗专家,尤其通晓苗族音乐和苗族古歌的演唱。杨雅伦的此次到来为与李老师录制苗族语音数据库做了前期准备。在杨雅伦的帮助下,李学德老师和其妻子田荣芬与ECura团队签署了同意录制语音数据库的纸质合同,之后又帮助他们在银行设立专门的账户以便接收合作款项。考虑到老年人不便长时间紧盯小小的手机屏幕,ECura团队为李老师的电脑配置了高品质麦克风以及便捷的录音软件。同时,雅伦和李学德一起整理了苗族古歌的词汇,让李老师更熟悉语料库录制的重点和流程。

此外,李学德老师还邀请杨雅伦和杨云惠前往他的家乡阿初鲁村,为雅伦介绍了当地音乐家,并参与了饭后的音乐表演。ECura团队与李老师的合作展示了团队对于搭建少数民族语音数据库的决心和行动。经过沟通和交流,李老师很好地理解了录制内容、方法并和团队形成良好关系。ECura团队也将继续致力于对苗族音乐文化的观察记录和保护

Bai Dabenqu Corpus Launched Through ECura Collaboration

On March 18, 2025, postdoctoral researcher Keyi Liu traveled to Dali, Yunnan Province, China, to meet with Sen Yang, a provincial-level representative inheritor of Bai Dabenqu in Yunnan. The visit aimed to advance collaboration on building a Bai Dabenqu corpus as part of the ECura project.

After Keyi Liu introduced the project’s objectives in detail to Sen Yang and Haizhen Guo, the two sides held an in-depth discussion on the specifics of the recording work. They specifically addressed aspects such as recording techniques, appropriate equipment selection, audio quality standards, and ensuring accuracy and completeness of bai vocabulary recordings. After discussions and a trial recording session, both parties officially signed the cooperation agreement on 24 March.

Sen Yang will record around 2,000 audio entries of commonly used Bai vocabulary, including Dabenqu-specific terms, their Chinese translations, and corresponding singing examples. He also presented the current progress of his ongoing work on transcribing and recording handwritten Dabenqu manuscripts, noting that 35 manuscripts are still awaiting further work, and expressed enthusiasm for future deeper collaboration with the ECura project.

2025年3月18日,博士后研究员刘可意前往中国云南大理,拜访白族大本曲传承人杨森,共同推进ECura项目的白族大本曲语料库建设。在刘可意向杨森和郭海珍 (杨森助理)详细介绍项目需求后,双方深入讨论了录制工作的具体细节,包括音频录制的方法、设备的选择、音质要求以及如何确保录制的词汇准确性和完整性。在前期讨论与试录工作基础上,双方于3月24日正式达成合作并签署协议。杨森将录制约2000条白族常用词汇音频,内容包括大本曲中的词汇及相应的中文翻译与演唱示例。此外,杨森还展示了当前正在进行的大本曲手抄本整理和录制进展,并指出目前共有35本手抄本需要继续整理与录制工作,期待未来与ECura项目开展更深入的合作。

 

Keyi Liu, Sen Yang and Haizhen Guo sign the agreement for the Bai Dabenqu corpus at Yang’s studio in Dali, 24 March 2025 (Photo provided by Keyi Liu)

Written and edited by Keyi Liu

28th CHIME Conference “Digital Futures for Chinese Music” supported by ECura at UCC

 

28th CHIME conference
28th CHIME Conference attendees (photo by Tianyu Liu)
Conference attendees at the Yaji (雅集) session (photo by Jacob Charles).
Jean-Luc Penso, the founder of Théâtre du Petit Miroir performed a puppet show ‘Fox Story’ inspired by traditional Chinese shadow puppets at the conference (photo by Tianyu Liu).

ECura Advisory Board Gathers to Review Progress and Plan Next Steps

ECura Project Advisory Board meeting in the Fleischmann Room of the Music Department at UCC (photo by Tianyu Liu)