兼职教授

当前位置:首页 >> 师资队伍 >> 兼职教授
  

庞在虎 兼职教授

20210618162400334741400312.png

庞在虎 博士,灵伴智能联合创始人,首席技术官(CTO)

北京石油化工学院人工智能研究院,兼职教授

 

个人简历

l            2006.09-2012.06 北京大学信息科学技术学院,博士研究生

l            2012.07-2014.06 北京大学信息科学技术学院,博士后,讲师

l            2014.06-至今    灵伴智能,联合创始人,首席技术官(CTO)

 

研究方向

语音信号处理、语音识别、语音合成、说话人识别、音频指纹分析与识别;

自然语言处理、语义理解、知识图谱;

图像文字识别、统计机器学习与深度学习;

智能语音交互系统、智能语音对话系统。

 

发表论文

1.Zaihu Pang, Shikui Tu, Xihong Wu, Lei Xu. A comparative study of RPCL and MCE based discriminative training methods for LVCSR. Neurocomputing,134,2014. (SCI/EI)

2.Zaihu Pang, Shikui Tu, Xihong Wu, Lei Xu. Discriminative GMM-HMM Acoustic Model Selection Using Two-level Bayesian Ying-Yang Harmony Learning. IScIDE 2012, Lecture Notes in Computer Science,7751 2013. (EI)

3.Zaihu Pang, Xihong Wu, Lei Xu. A comparative study of RPCL and MCE based discriminative training methods for LVCSR. IScIDE 2011, Lecture Notes in Computer Science, 7202, pp.27-34, 2012. (EI)

4.Zaihu Pang, Xihong Wu, Lei Xu. Discriminative training of GMM-HMM acoustic model by RPCL type Bayesian Ying-Yang harmony learning. Sino-foreign-interchange Workshop on Intelligence Science and Intelligent Data Engineering, Xi’an, China, October 23-26, 2011.

5.Zaihu Pang, Shikui Tu, Dan Su, Xihong Wu, Lei Xu. Discriminative training of GMM-HMM acoustic model by RPCL learning. A special issue on Machine learning and intelligence science: IScIDE2010 (B), Frontiers of Electrical and Electronic Engineering in China, 6(2), pp.283-290, 2011.

6.Zaihu Pang, Dan Su, Xihong Wu, Lei Xu. Rival penalized competitive learning based discriminative training of acoustic model for LVCSR. Sino-foreign-interchange Workshop on Intelligence Science and Intelligent Data Engineering, Harbin, China, p.14, June 3-5, 2010.

7.Xiangang Li, Zaihu Pang, Xihong Wu. Lightly Supervised Acoustic Model Training for Mandarin Continuous Speech Recognition. IScIDE 2012, Lecture Notes in Computer Science,7751, 2013. (EI)

8.Xiangang Li, Dan Su, Zaihu Pang, Xihong Wu. Probabilistic speaker-class based acoustic modeling for large vocabulary continuous speech recognition, In Proceedings of INTERSPEECH’2012, Portland, Oregon, USA, September 9-13, 2012. (EI)

9.李先刚, 张晨炜, 庞在虎, 吴玺宏.大词汇量连续语音识别解码器优化研究与实现[C], 第十二届全国人机语音通讯学术会议(NCMMSC'2013) , 2013.

 

发明专利

1. 一种针对短语音的说话人识别模型的训练方法及装置(专利号:ZL 201710651305.2)

2. 一种虚拟人物语音与口型同步的建模与控制方法(专利号:ZL 201810199537.3)

3. 应用于自动翻页扫描仪的吸书机构及吸书方法(专利号:ZL 201910732550.5)

4. 一种面向智能语音对话系统的多层级对话分析方法(专利号:ZL 201711130800.5)

5. 一种面向非自由谈话式汉语口语的对话行为分析方法(专利号:ZL 201711087012.2)

6. 一种说话人自动标注方法(专利号:ZL 201810673628.6)

7. 一种电话录音未接通原因识别方法(专利号:ZL 201910076136.3)

8. 数据加密的客服系统及客服方法(专利号:ZL 202010556891.4)

9. 一种文档图像的色彩模式自动设定方法(专利号:ZL 202010599925.8)

10. 电话客服系统及电话客服弹性转接方法(专利号:ZL 202010659558.6)

11. 声音优化方法及声音优化系统(专利号:ZL 202010825664.7)

12. 人机混排的客服方法及系统(专利号:ZL 202010125106.X)

13. 人机协同的电话客服方法及系统(专利号:ZL 202010125647.2)