HTHiroko Terasawa

Hiroko Terasawa

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Hiroko Terasawa has a total of 8 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Wavetable Synthesis Using CVAE for Timbre Control Based on Semantic Label

    by Tsugumasa Yutani, Yuya Yamamoto, Shuyo Nakatani, Hiroko Terasawa

    Synthesizers are essential in modern music production. However, their complex timbre parameters, often filled with technical terms, require expertise. This research introduces a method of timbre control in wavetable synthesis that is intuitive and sensible and utilizes semantic labels. Using a conditional variational autoencoder (CVAE), users can select a wavetable and define the timbre with labels such as bright, warm, and rich. The CVAE model, featuring convolutional and upsampling layers, effectively captures the wavetable nuances, ensuring real-time performance owing to their processing in the time domain. Experiments demonstrate that this approach allows for real-time, effective control of the timbre of the wavetable using semantic inputs and aims for intuitive timbre control through data-based semantic control.

  • Open Access

    PrimaDNN': A Characteristics-aware DNN Customization for Singing Technique Detection

    by Yuya Yamamoto, Juhan Nam, Hiroko Terasawa

    Professional vocalists modulate their voice timbre or pitch to make their vocal performance more expressive. Such fluctuations are called singing techniques. Automatic detection of singing techniques from audio tracks can be beneficial to understand how each singer expresses the performance, yet it can also be difficult due to the wide variety of the singing techniques. A deep neural network (DNN) model can handle such variety; however, there might be a possibility that considering the characteristics of the data improves the performance of singing technique detection. In this paper, we propose PrimaDNN, a CRNN model with a characteristics-oriented improvement. The features of the model are: 1) input feature representation based on auxiliary pitch information and multi-resolution mel spectrograms, 2) Convolution module based on the Squeeze-and-excitation (SENet) and the Instance normalization. In the results of J-POP singing technique detection, PrimaDNN achieved the best results of 44.9% at the overall macro-F measure, compared to conventional works. We also found that the contribution of each component varies depending on the type of singing technique.

  • Open Access

    Analysis and Detection of Singing Techniques in Repertoires of J-POP Solo Singers

    by Yuya Yamamoto, Juhan Nam, Hiroko Terasawa

    In this paper, we focus on singing techniques within the scope of music information retrieval research. We investigate how singers use singing techniques using real-world recordings of famous solo singers in Japanese popular music songs (J-POP). First, we built a new dataset of singing techniques. The dataset consists of 168 commercial J-POP songs, and each song is annotated using various singing techniques with timestamps and vocal pitch contours. We also present descriptive statistics of singing techniques on the dataset to clarify what and how often singing techniques appear. We further explored the difficulty of the automatic detection of singing techniques using previously proposed machine learning techniques. In the detection, we also investigate the effectiveness of auxiliary information (i.e., pitch and distribution of label duration), not only providing the baseline. The best result achieves 40.4% at macro-average F-measure on nine-way multi-class detection. We provide the annotation of the dataset and its detail on the appendix website 0 .


  • Open Access

    Music in Quarantine: Connections Between Changes in Lifestyle, Psychological States, and Musical Behaviors During COVID-19 Pandemic

    by Hiroko Terasawa, Masaki Matsubara, Visda Goudarzi, Makiko Sadakata

    Music is not only the art of organized sound but also a compound of social interaction among people, built upon social and environmental foundations. Since the beginning of the COVID-19 outbreak, containment measures such as shelter-in-place, lockdown, social distancing, and self-quarantine have severely impacted the foundation of human society, resulting in a drastic change in our everyday experience. In this paper, the relationships between musical behavior, lifestyle, and psychological states during the shelter-in-place period of the COVID-19 pandemic are investigated. An online survey on musical experience, lifestyle changes, stress level, musical behaviors, media usage, and environmental sound perception was conducted. The survey was conducted in early June 2020. Responses from 620 people in 24 countries were collected, with the large proportion of the responses coming from the U.S. (55.5%) and India (21.4%). Structural equation modeling (SEM) analysis revealed causal relationships between lifestyle, stress, and music behaviors. Elements such as stress-level change, work risk, and staying home contribute to changes in musical experiences, such as moderating emotion with music, feeling emotional with music, and being more attentive to music. Stress-level change was correlated with work risk and income change, and people who started living with others due to the outbreak, especially with their children, indicated less change in stress level. People with more stress-level change tended to use music more purposefully for their mental well-being, such as to moderate emotions, to influence mood, and to relax. In addition, people with more stress-level change tend to be more annoyed by neighbors' noise. Housing type was not directly associated with annoyance; however, attention to environmental sounds decreased when the housing type was smaller. Attention to environmental and musical sounds and the emotional responses to them are highly inter-correlated. Multi-group SEM based on musicians showed that the causal relationship structure for professional musicians differs from that of less-experienced musicians. For professional musicians, staying at home was the only component that caused all musical behavior changes; stress did not cause musical behavior changes. Regarding Internet use, listening to music via YouTube and streaming was preferred over TV and radio, especially among less-experienced musicians, while participation in the online music community was preferred by more advanced musicians. This work suggests that social, environmental, and personal factors and limitations influence the changes in our musical behavior, perception of sonic experience, and emotional recognition, and that people actively accommodated the unusual pandemic situations using music and Internet technologies.