Picture for Zhengkun Tian

Zhengkun Tian

Peak-First CTC: Reducing the Peak Latency of CTC Models by Applying Peak-First Regularization

Nov 07, 2022
Zhengkun Tian, Hongyu Xiang, Min Li, Feifei Lin, Ke Ding, Guanglu Wan

Figure 1 for Peak-First CTC: Reducing the Peak Latency of CTC Models by Applying Peak-First Regularization

Figure 2 for Peak-First CTC: Reducing the Peak Latency of CTC Models by Applying Peak-First Regularization

Figure 3 for Peak-First CTC: Reducing the Peak Latency of CTC Models by Applying Peak-First Regularization

* Submitted to ICASSP 2023(5 pages, 2 figures)

Access Paper or Ask Questions

System Fingerprints Detection for DeepFake Audio: An Initial Dataset and Investigation

Aug 21, 2022
Xinrui Yan, Jiangyan Yi, Jianhua Tao, Chenglong Wang, Haoxin Ma, Zhengkun Tian, Ruibo Fu

Figure 1 for System Fingerprints Detection for DeepFake Audio: An Initial Dataset and Investigation

Figure 2 for System Fingerprints Detection for DeepFake Audio: An Initial Dataset and Investigation

Figure 3 for System Fingerprints Detection for DeepFake Audio: An Initial Dataset and Investigation

Figure 4 for System Fingerprints Detection for DeepFake Audio: An Initial Dataset and Investigation

* 9 pages, 3 figures. arXiv admin note: text overlap with arXiv:2208.09646

Access Paper or Ask Questions

ADD 2022: the First Audio Deep Synthesis Detection Challenge

Feb 26, 2022
Jiangyan Yi, Ruibo Fu, Jianhua Tao, Shuai Nie, Haoxin Ma, Chenglong Wang, Tao Wang, Zhengkun Tian, Ye Bai, Cunhang Fan, Shan Liang, Shiming Wang, Shuai Zhang, Xinrui Yan, Le Xu, Zhengqi Wen, Haizhou Li, Zheng Lian, Bin Liu

Figure 1 for ADD 2022: the First Audio Deep Synthesis Detection Challenge

Figure 2 for ADD 2022: the First Audio Deep Synthesis Detection Challenge

Figure 3 for ADD 2022: the First Audio Deep Synthesis Detection Challenge

Figure 4 for ADD 2022: the First Audio Deep Synthesis Detection Challenge

* Accepted by ICASSP 2022

Access Paper or Ask Questions

Reducing language context confusion for end-to-end code-switching automatic speech recognition

Jan 28, 2022
Shuai Zhang, Jiangyan Yi, Zhengkun Tian, Jianhua Tao, Yu Ting Yeung, Liqun Deng

Figure 1 for Reducing language context confusion for end-to-end code-switching automatic speech recognition

Figure 2 for Reducing language context confusion for end-to-end code-switching automatic speech recognition

Figure 3 for Reducing language context confusion for end-to-end code-switching automatic speech recognition

Figure 4 for Reducing language context confusion for end-to-end code-switching automatic speech recognition

* arXiv admin note: text overlap with arXiv:2010.14798

Access Paper or Ask Questions

Continual Learning for Fake Audio Detection

Apr 15, 2021
Haoxin Ma, Jiangyan Yi, Jianhua Tao, Ye Bai, Zhengkun Tian, Chenglong Wang

Figure 1 for Continual Learning for Fake Audio Detection

Figure 2 for Continual Learning for Fake Audio Detection

Figure 3 for Continual Learning for Fake Audio Detection

Figure 4 for Continual Learning for Fake Audio Detection

* 5 pages, conference

Access Paper or Ask Questions

Half-Truth: A Partially Fake Audio Detection Dataset

Apr 08, 2021
Jiangyan Yi, Ye Bai, Jianhua Tao, Zhengkun Tian, Chenglong Wang, Tao Wang, Ruibo Fu

Figure 1 for Half-Truth: A Partially Fake Audio Detection Dataset

Figure 2 for Half-Truth: A Partially Fake Audio Detection Dataset

Figure 3 for Half-Truth: A Partially Fake Audio Detection Dataset

Figure 4 for Half-Truth: A Partially Fake Audio Detection Dataset

* submitted to Interspeech 2021

Access Paper or Ask Questions

FSR: Accelerating the Inference Process of Transducer-Based Models by Applying Fast-Skip Regularization

Apr 07, 2021
Zhengkun Tian, Jiangyan Yi, Ye Bai, Jianhua Tao, Shuai Zhang, Zhengqi Wen

Figure 1 for FSR: Accelerating the Inference Process of Transducer-Based Models by Applying Fast-Skip Regularization

Figure 2 for FSR: Accelerating the Inference Process of Transducer-Based Models by Applying Fast-Skip Regularization

Figure 3 for FSR: Accelerating the Inference Process of Transducer-Based Models by Applying Fast-Skip Regularization

Figure 4 for FSR: Accelerating the Inference Process of Transducer-Based Models by Applying Fast-Skip Regularization

* Submitted to INTERSPEECH2021

Access Paper or Ask Questions