| Description |
1 online resource (143 pages) |
| Series |
Lecture Notes in Electrical Engineering Ser. ; v. 635 |
|
Lecture notes in electrical engineering ; v. 635.
|
| Contents |
Intro -- Preface -- Contents -- Part I Speech Processing -- 1 Bandwidth Extension WaveNet for Bone-Conducted Speech Enhancement -- 1 Introduction -- 2 The Proposed Method -- 2.1 The Overall Framework -- 2.2 Bandwidth Extension WaveNet -- 2.2.1 Up-sampling Module of Cross Sampling Rate. -- 2.2.2 Blocks of Dilated Convolution. -- 2.2.2 Discretized Logistic Mixture Distribution. -- 3 Experiment -- 3.1 Dataset and Evaluation Metrics -- 3.2 Experimental Setup and Baseline System -- 3.2.1 Experimental Setup. -- 3.2.2 Baseline System. -- 3.3 Results and Analysis. -- 4 Conclusion -- 5 Acknowledgments |
|
3.2 Evaluation of operation order detection -- 4 Conclusion -- 5 Acknowledgments -- References -- Part II Music Information Retrieval -- 4 Singing Voice Detection Using Multi-Feature Deep Fusion with CNN -- 1 Introduction -- 2 Proposed SVD system -- 2.1 Singing Voice Separation -- 2.2 Feature Extraction -- 2.3 CNN for feature deep fusion and Classification -- 2.4 Post Process -- 3 Experiments and results -- 3.1 Benchmark Datasets -- 3.2 Evaluation -- 3.3 Results -- 4 Conclusions -- 5 Acknowledgement -- References -- 5 A Multi-task Learning Approach for Melody Extraction -- 1 Introduction |
|
2 Related Works -- 2.1 Deep Learning for Melody Extraction -- 2.2 Multi-task Learning -- 3 Proposed Method -- 3.1 Architecture -- 3.2 Multi-task Loss -- 3.3 Training -- 3.4 Inferencing -- 4 Experiments -- 4.1 Datasets -- 4.2 Metrics -- 4.3 Effects of Auxiliary Weight for Multi-task Learning -- 4.4 Comparison with State-of-the-Art Methods -- 5 Conclusions and Discussions -- 6 Acknowledgements -- References -- 6 A post-processing of onset detection based on verification with neural network -- 1 Introduction -- 1.1 Common onset detection -- 1.2 Novelty onset detection -- 1.3 Our works |
|
2 Onset detection -- 2.1 Onset function -- 2.2 Peak-picking algorithm -- 2.3 Verification -- 3 Evaluation -- 3.1 Dataset -- 3.2 Classification and Onset Detection -- 4 Result -- 4.1 Peak-picking in sliding window -- 4.2 Baseline and Verification -- 4.3 Result and Discussion -- 5 Conclusion -- References -- 7 Transfer learning for music classification and regression tasks using artist tags -- 1 Introduction -- 2 Proposed method -- 2.1 Preprocess of audio -- 2.2 Model for artist tagging -- 2.3 Feature transfer and extraction -- 2.4 Classification -- 3 Datasets -- 3.1 Source dataset |
| Summary |
The book presents selected papers that have been accepted at the seventh Conference on Sound and Music Technology (CSMT) in December 2019, held in Harbin, Hei Long Jiang, China. CSMT is a domestic conference focusing on audio processing and understanding with bias on music and acoustic signals. The primary aim of the conference is to promote the collaboration between art society and technical society in China. The organisers of CSMT hope the conference can serve as a platform for interdisciplinary research. In this proceeding, the paper included covers a wide range topic from speech, signal processing and music understanding, which demonstrates the target of CSMT merging arts and science research together |
| Bibliography |
References-2 Naturalness evaluation of synthetic speech based on residual learning networks-1 Introduction-2 Methods-2.1 Database-2.2 Features Extraction-2.3 The ResNet Based Speech Evaluation System-3 Experiments-4 Conclusions And Future Work-Acknowledgement-References-3 Detection of Operation Type and Order for Digital Speech-1 Introduction-2 Proposed detection algorithm-2.1 Effect of various operations on spectrograms-2.2 Feature extraction-2.3 Classification-3 Experimental results-3.1 Effect of various operations on spectrograms |
| Notes |
3.2 Target datasets |
|
Print version record |
| Subject |
Computer music -- Congresses
|
|
Computer music
|
| Genre/Form |
proceedings (reports)
|
|
Conference papers and proceedings
|
|
Conference papers and proceedings.
|
|
Actes de congrès.
|
| Form |
Electronic book
|
| Author |
Li, Haifeng (Computer science professor)
|
|
Li, Shengchen
|
|
Ma, Lin
|
|
Fang, Chunying
|
|
Zhu, Yidan
|
| ISBN |
9789811527562 |
|
9811527563 |
|
