Papers by Takaaki Saeki
Personalized Filled-pause Generation with Group-wise Prediction Models (2022.lrec-1)
Copied to clipboard
| Challenge: | Disfluency generation is a method to generate personalized filled pauses (FPs) compared with fluent text generation, it is difficult to predict them because of the sparsity of position and frequency difference between more and less frequently used FPs. |
| Approach: | They propose a method to generate personalized filled pauses (FPs) by group-wise prediction models. |
| Outcome: | The proposed method generates personalized filled pauses (FPs) with group-wise prediction models. |
SSR7000: A Synchronized Corpus of Ultrasound Tongue Imaging for End-to-End Silent Speech Recognition (2022.lrec-1)
Copied to clipboard
| Challenge: | Existing models for silent speech recognition have not been able to capture large datasets. |
| Approach: | They present a corpus of synchronized ultrasound tongue and lip images for silent speech recognition . they use a large dataset to exploit the performance of the end-to-end models . |
| Outcome: | The proposed model outperforms existing models in a large dataset of ultrasound tongue and lip images . the dataset contains more utterances per person than any other based on ultrasound imaging . |