no code implementations • LREC 2022 • Sheng Li, Jiyi Li, Qianying Liu, Zhuo Gong
Moreover, based on the speech collection, we proposed a neural network-based frame-by-frame mapping method to recover the speech content by converting from the adversarial speech to the human speech.
no code implementations • SIGDIAL (ACL) 2022 • Longfei Yang, Jiyi Li, Sheng Li, Takahiro Shinozaki
In the slot self-attention layers, we force each slot to involve information from the other k prominent slots and mask the rest out.
Dialogue State Tracking Multi-domain Dialogue State Tracking +1
no code implementations • EMNLP 2020 • Wenshuo Yang, Jiyi Li, Fumiyo Fukumoto, Yanming Ye
The data imbalance problem is a crucial issue for the multi-label text classification.
Ranked #11 on Spectral Reconstruction on ARAD-1K
no code implementations • EMNLP (sdp) 2020 • Jiyi Li, Ayaka Sato, Kazuya Shimura, Fumiyo Fukumoto
To handle the small size of published datasets on the target aspect of scores, we propose a multi-task approach to leverage additional information from other aspects of scores for improving the performance of the target.
1 code implementation • 10 Apr 2024 • Xinfeng Wang, Fumiyo Fukumoto, Jin Cui, Yoshimi Suzuki, Jiyi Li, Dongjin Yu
To tackle the skewed distribution, we propose two strategies for disentangling interactions: (1) modeling individual biases to learn unbiased item embeddings, and (2) incorporating item popularity with positional encoding.
no code implementations • 15 Mar 2024 • Jin Cui, Fumiyo Fukumoto, Xinfeng Wang, Yoshimi Suzuki, Jiyi Li, Noriko Tomuro, Wanzeng Kong
To address the issue of multiple aspect categories and sentiment entanglement, we propose a hierarchical disentanglement module to extract distinct categories and sentiment features.
no code implementations • 18 Jan 2024 • Jiyi Li
Whether Large Language Models (LLMs) can outperform crowdsourcing on the data annotation task is attracting interest recently.
no code implementations • 21 Nov 2023 • Wenqing Wei, Zhengdong Yang, Yuan Gao, Jiyi Li, Chenhui Chu, Shogo Okada, Sheng Li
The early-stage Alzheimer's disease (AD) detection has been considered an important field of medical studies.
no code implementations • 17 Nov 2023 • Xiaojiao Chen, Sheng Li, Jiyi Li, Hao Huang, Yang Cao, Liang He
This paper demonstrates that an attacker can extract speaker information by querying speaker-adapted speech recognition (ASR) systems.
no code implementations • 17 Nov 2023 • Xiaojiao Chen, Sheng Li, Jiyi Li, Hao Huang, Yang Cao, Liang He
Current speaker anonymization methods, especially with self-supervised learning (SSL) models, require massive computational resources when hiding speaker identity.
1 code implementation • 8 Feb 2023 • Xiaotian Lu, Jiyi Li, Koh Takeuchi, Hisashi Kashima
Crowdsourcing has been used to collect data at scale in numerous fields.
1 code implementation • EMNLP 2021 • Zhiwei Zhang, Jiyi Li, Fumiyo Fukumoto, Yanming Ye
In addition, we enhance the information exchanges and constraints among tasks by proposing a regularization term between the sentence attention scores of abstract retrieval and the estimated outputs of rational selection.
no code implementations • 29 Jun 2021 • Jingzheng Li, Hailong Sun, Jiyi Li, Zhijun Chen, Renshuai Tao, Yufei Ge
Learning from multiple annotators aims to induce a high-quality classifier from training instances, where each of them is associated with a set of possibly noisy labels provided by multiple annotators under the influence of their varying abilities and own biases.
no code implementations • COLING 2020 • Panitan Muangkammuen, Sheng Xu, Fumiyo Fukumoto, Kanda Runapongsa Saikaew, Jiyi Li
Local coherence relation between two phrases/sentences such as cause-effect and contrast gives a strong influence of whether a text is well-structured or not.
no code implementations • 1 Aug 2020 • Yukino Baba, Jiyi Li, Hisashi Kashima
We propose an approach, called CrowDEA, which estimates the embeddings of the ideas in the multiple-criteria preference space, the best viewpoint for each idea, and preference criterion for each evaluator, to obtain a set of frontier ideas.
no code implementations • WS 2020 • Chu Su, ong, Fumiyo Fukumoto, Xiaoxi Huang, Jiyi Li, Rongbo Wang, Zhiqun Chen
Machine metaphor understanding is one of the major topics in NLP.
no code implementations • WS 2019 • Jiyi Li, Fumiyo Fukumoto
To ensure the quality of the crowdsourced data, people can assign multiple workers to one question and then aggregate the multiple answers with diverse quality into a golden one.
1 code implementation • ACL 2019 • Kazuya Shimura, Jiyi Li, Fumiyo Fukumoto
Distributions of the senses of words are often highly skewed and give a strong influence of the domain of a document.
1 code implementation • EMNLP 2018 • Kazuya Shimura, Jiyi Li, Fumiyo Fukumoto
The lower the HS level, the less the categorization performance.