English
Related papers

Related papers: MT3: Multi-Task Multitrack Music Transcription

200 papers

A key challenge in machine learning is to generalize from training data to an application domain of interest. This work generalizes the recently-proposed mixture invariant training (MixIT) algorithm to perform unsupervised learning in the…

Sound · Computer Science 2024-03-25 Cong Han , Kevin Wilson , Scott Wisdom , John R. Hershey

In this progress paper the previous results of the single note recognition by deep learning are presented. The several ways for data augmentation and "artificial semantic" augmentation are proposed to enhance efficiency of deep learning…

As an important component of multimedia analysis tasks, audio classification aims to discriminate between different audio signal types and has received intensive attention due to its wide applications. Generally speaking, the raw signal can…

Multimedia · Computer Science 2020-02-25 Liang Gao , Kele Xu , Huaimin Wang , Yuxing Peng

Speech classification tasks often require powerful language understanding models to grasp useful features, which becomes problematic when limited training data is available. To attain superior classification performance, we propose to…

Computation and Language · Computer Science 2024-07-26 Nicolae-Catalin Ristea , Andrei Anghel , Radu Tudor Ionescu

Music transcription, which deals with the conversion of music sources into a structured digital format, is a key problem for Music Information Retrieval (MIR). When addressing this challenge in computational terms, the MIR community follows…

Multimedia · Computer Science 2022-08-29 María Alfaro-Contreras , Jose J. Valero-Mas , José M. Iñesta , Jorge Calvo-Zaragoza

Unsupervised cross-lingual pretraining has achieved strong results in neural machine translation (NMT), by drastically reducing the need for large parallel data. Most approaches adapt masked-language modeling (MLM) to sequence-to-sequence…

Computation and Language · Computer Science 2021-06-11 Christos Baziotis , Ivan Titov , Alexandra Birch , Barry Haddow

Mood recognition is an important problem in music informatics and has key applications in music discovery and recommendation. These applications have become even more relevant with the rise of music streaming. Our work investigates the…

Sound · Computer Science 2021-10-12 Rajnish Kumar , Manjeet Dahiya

Applications of deep learning to automatic multitrack mixing are largely unexplored. This is partly due to the limited available data, coupled with the fact that such data is relatively unstructured and variable. To address these…

Audio and Speech Processing · Electrical Eng. & Systems 2020-10-21 Christian J. Steinmetz , Jordi Pons , Santiago Pascual , Joan Serrà

We propose a novel symbolic music representation and Generative Adversarial Network (GAN) framework specially designed for symbolic multitrack music generation. The main theme of symbolic music generation primarily encompasses the…

Sound · Computer Science 2024-09-04 Jinlong Zhu , Keigo Sakurai , Ren Togo , Takahiro Ogawa , Miki Haseyama

Self-supervised Audio Transformers (SAT) enable great success in many downstream speech applications like ASR, but how they work has not been widely explored yet. In this work, we present multiple strategies for the analysis of attention…

Computation and Language · Computer Science 2020-08-12 Shu-wen Yang , Andy T. Liu , Hung-yi Lee

Conditional music generation offers significant advantages in terms of user convenience and control, presenting great potential in AI-generated content research. However, building conditional generative systems for multitrack popular songs…

Sound · Computer Science 2025-10-27 Jing Luo , Xinyu Yang , Dorien Herremans

Bootstrap-based Self-Supervised Learning (SSL) has achieved remarkable progress in audio understanding. However, existing methods typically operate at a single level of granularity, limiting their ability to model the diverse temporal and…

Audio and Speech Processing · Electrical Eng. & Systems 2026-01-30 Bing Han , Chushu Zhou , Yifan Yang , Wei Wang , Chenda Li , Wangyou Zhang , Yanmin Qian

Traditional methods to tackle many music information retrieval tasks typically follow a two-step architecture: feature engineering followed by a simple learning algorithm. In these "shallow" architectures, feature engineering and learning…

Sound · Computer Science 2015-11-18 Peter Li , Jiyuan Qian , Tian Wang

Human language is often multimodal, which comprehends a mixture of natural language, facial gestures, and acoustic behaviors. However, two major challenges in modeling such multimodal human language time-series data exist: 1) inherent data…

Computation and Language · Computer Science 2019-06-04 Yao-Hung Hubert Tsai , Shaojie Bai , Paul Pu Liang , J. Zico Kolter , Louis-Philippe Morency , Ruslan Salakhutdinov

Extracting pitch information from music recordings is a challenging but important problem in music signal processing. Frame-wise transcription or multi-pitch estimation aims for detecting the simultaneous activity of pitches in polyphonic…

Sound · Computer Science 2022-02-21 Christof Weiß , Geoffroy Peeters

The advent of hyper-scale and general-purpose pre-trained models is shifting the paradigm of building task-specific models for target tasks. In the field of audio research, task-agnostic pre-trained models with high transferability and…

Audio and Speech Processing · Electrical Eng. & Systems 2023-03-03 Ju-ho Kim , Jungwoo Heo , Hyun-seo Shin , Chan-yeong Lim , Ha-Jin Yu

Accurate detection of disfluencies in spoken language is crucial for enhancing the performance of automatic speech and language processing systems, as well as fostering the development of more inclusive speech and language technologies.…

Sound · Computer Science 2025-06-24 Duygu Altinok

Multi-source translation systems translate from multiple languages to a single target language. By using information from these multiple sources, these systems achieve large gains in accuracy. To train these systems, it is necessary to have…

Computation and Language · Computer Science 2018-11-09 Yuta Nishimura , Katsuhito Sudoh , Graham Neubig , Satoshi Nakamura

A central goal in automatic music transcription is to detect individual note events in music recordings. An important variant is instrument-dependent music transcription where methods can use calibration data for the instruments in use.…

Sound · Computer Science 2017-11-01 Sebastian Ewert , Mark B. Sandler

Multimodal machine translation (MMT) aims to improve translation quality by incorporating information from other modalities, such as vision. Previous MMT systems mainly focus on better access and use of visual information and tend to…

Computation and Language · Computer Science 2023-09-06 Yaoming Zhu , Zewei Sun , Shanbo Cheng , Luyang Huang , Liwei Wu , Mingxuan Wang