[Adapt] FW: 【学术报告】字节跳动AI Lab孔秋强 - 音频分类与检测 AI for Sound: Large-scale audio tagging with deep neural networks

[Kenny Zhu]

@zelin

-----Original Message-----
From: Mengyue Wu [mailto:mengyuewu at sjtu.edu.cn] 
Sent: 2021年1月11日 16:25
To: all at cs.sjtu.edu.cn
Subject: 【学术报告】字节跳动AI Lab孔秋强 - 音频分类与检测 AI for Sound: Large-scale audio tagging with deep neural networks

各位老师好：

 现邀请到了来自字节跳动AI Lab的研究员孔秋强博士来给题为“AI for Sound: Large-scale audio tagging with deep neural networks”的学术报告。欢迎您和您的学生参加！

 具体信息如下：

 时间: 1月13日（本周三）10:30-11:30
 地点: 电信群楼3-404会议室
 主持: 吴梦玥


Title: AI for Sound: Large-scale audio tagging with deep neural networks

Abstract: 音频分类、检测是音频信号处理领域的重要课题。音频信号处理涵盖了音频打标签、音频场景分类、音乐分类、语音端点检测、异常事件检测等。近些年，基于神经网络的方法在音频信号处理中获得了许多成功的应用，并在结果上超过了传统信号处理的方法。该讲义介绍了音频分类、检测的近些年的发展历史，以及以神经网络为基础的一系列音频检测与分类方法。大部分音频信号处理的工作使用较小的数据集，限制了分类系统的性能。该讲义介绍了最新提出的在大规模音频数据库AudioSet上训练通用的音频分类模型，可以实时检测自然界中527种类型的声音。系统取得了目前截至最高的平均准确率mAP 0.439。为了探究该系统的泛化性能，该预训练模型被应用在8个音频分类、检测数据上，取得了比无预训练模型更好的效果。该讲义总结并展望了音频信号处理的未来研究方向。

Bio: Qiuqiang Kong received his Ph.D. degree from University of Surrey, Guildford, UK in 2020. Following his PhD, he joined ByteDance AI Lab as a research scientist. His research topic includes the classification, detection and separation of general sounds and music. He is known for developing attention neural networks for audio tagging, and winning the audio tagging task in the detection and classification of acoustic scenes and events (DCASE) challenge in 2017. He has authored paper in journals and conferences including IEEE/ACM Transactions on Audio, Speech, and Language Processing (TASLP), ICASSP, INTERSPEECH, IJCAI, DCASE, EUSIPCO, LVA-ICA, etc. He has been cited 773 times, with an H-index 15 till Aug. 2020. He was nominated as the postgraduate research student of the year in University of Surrey, 2019. He is a frequent reviewer for over ten world well known journals and conferences including TASLP, TMM, SPL, TKDD, JASM, EURASIP, Neurocomputin, Neural Networks, ISMIR, CSMT, etc.

Personal Website: https://qiuqiangkong.github.io/


谢谢！
吴梦玥