Triplet Convolutional Network for Music Version Identification

Qi, Xiaoyu; Yang, Dingcheng; Chen, Xiaoou

doi:10.1007/978-3-319-73603-7_44

Cited by 8 publications

(8 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We conducted comparative experiments to evaluate the effectiveness of our method. We took a ranking-based evaluation approach [25], [47]- [49] to quantitatively evaluate the performance of compatibility estimation. Ranking-based evaluation is used to assess how accurately a target can be found from multiple candidates.…”

Section: Discussionmentioning

confidence: 99%

“…Because data of different modalities can be treated as identical data in a jointembedding space and trained under a common metric, deep metric learning and joint-embedding techniques perform well together. In MIR-related tasks, deep metric learning succeeds in learning joint representations over several modalities such as a vocal and mix [23], vocal imitation and sound recording [24], [25], animal sounds [26], sheet music and audio spectrograms [27], music and image [28]- [31], and music and video [21], [22]. The target pair for the metric learning described in this paper consists of a vocal track and an accompaniment track.…”

Section: B Self-supervised and Joint-embedding Techniquesmentioning

confidence: 99%

“…We compared four conditions: a baseline method based on a chromagram with (1) cosine similarity (Cos-Chromagram) and (2) dot product (Dot-Chromagram), and our method with (3) cosine similarity (Cos) and (4) dot product (Dot). The use of a chromagram as a feature vector can be considered a typical rule-based approach, and previous studies took this approach in combination with either cosine similarity or dot product [2], [4], [25], [51], [52]. This is why we included conditions (1) and (2) as the baseline methods.…”

Section: Conditionsmentioning

confidence: 99%

See 2 more Smart Citations

Vocal-Accompaniment Compatibility Estimation Using Self-Supervised and Joint-Embedding Techniques

et al. 2021

View full text Add to dashboard Cite

We propose a learning-based method of estimating the compatibility between vocal and accompaniment audio tracks, i.e., how well they go with each other when played simultaneously. This task is challenging because it is difficult to formulate hand-crafted rules or construct a large labeled dataset to perform supervised learning. Our method uses self-supervised and joint-embedding techniques for estimating vocal-accompaniment compatibility. We train vocal and accompaniment encoders to learn a jointembedding space of vocal and accompaniment tracks, where the embedded feature vectors of a compatible pair of vocal and accompaniment tracks lie close to each other and those of an incompatible pair lie far from each other. To address the lack of large labeled datasets consisting of compatible and incompatible pairs of vocal and accompaniment tracks, we propose generating such a dataset from songs using singing voice separation techniques, with which songs are separated into pairs of vocal and accompaniment tracks, and then original pairs are assumed to be compatible, and other random pairs are not. We achieved this training by constructing a large dataset containing 910,803 songs and evaluated the effectiveness of our method using ranking-based evaluation methods.INDEX TERMS Vocal-accompaniment compatibility, metric learning, music signal processing, music information retrieval.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: B Self-supervised and Joint-embedding Techniquesmentioning

confidence: 99%

Section: Conditionsmentioning

confidence: 99%

See 1 more Smart Citation

Vocal-Accompaniment Compatibility Estimation Using Self-Supervised and Joint-Embedding Techniques

et al. 2021

View full text Add to dashboard Cite

show abstract

“…At the lower end of the specificity scale are tasks such as genre recognition [36]. A medium-level specificity is associated with tasks such as audio matching [12,14], version identification [31,37], live song detection [28,38], and cover song retrieval [8,10,13,15,16,[21][22][23][24][25][26][27]29,30]. In all of these tasks, one allows for variations as they typically occur in different performances and arrangements of a piece of music.…”

Section: Related Workmentioning

confidence: 99%

“…An example of such a loss function is the hinge loss, which has been used to learn representations for cross-modal score-to-audio embeddings [45] and for artist similarity [46]. A related loss function is the triplet loss, which was developed for the task of face recognition [33] and then adapted for audio and music processing tasks such as speech retrieval [47], sound event classification [48], audio fingerprinting [49], artist clustering [50], music similarity [51], and cover song retrieval [37]. The latter study is conceptually similar to our approach, but presents results that are worse than those achieved by more traditional approaches [21].…”

Section: Neural Network With Triplet Lossmentioning

confidence: 99%

Learning Low-Dimensional Embeddings of Audio Shingles for Cross-Version Retrieval of Classical Music

Zalkow

Müller

2019

Applied Sciences

View full text Add to dashboard Cite

Cross-version music retrieval aims at identifying all versions of a given piece of music using a short query audio fragment. One previous approach, which is particularly suited for Western classical music, is based on a nearest neighbor search using short sequences of chroma features, also referred to as audio shingles. From the viewpoint of efficiency, indexing and dimensionality reduction are important aspects. In this paper, we extend previous work by adapting two embedding techniques; one is based on classical principle component analysis, and the other is based on neural networks with triplet loss. Furthermore, we report on systematically conducted experiments with Western classical music recordings and discuss the trade-off between retrieval quality and embedding dimensionality. As one main result, we show that, using neural networks, one can reduce the audio shingles from 240 to fewer than 8 dimensions with only a moderate loss in retrieval accuracy. In addition, we present extended experiments with databases of different sizes and different query lengths to test the scalability and generalizability of the dimensionality reduction methods. We also provide a more detailed view into the retrieval problem by analyzing the distances that appear in the nearest neighbor search.

show abstract

Accurate and Scalable Version Identification Using Musically-Motivated Embeddings

Yesiler

Serrà²

2020

ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

The version identification (VI) task deals with the automatic detection of recordings that correspond to the same underlying musical piece. Despite many efforts, VI is still an open problem, with much room for improvement, specially with regard to combining accuracy and scalability. In this paper, we present MOVE, a musically-motivated method for accurate and scalable version identification. MOVE achieves state-of-the-art performance on two publicly-available benchmark sets by learning scalable embeddings in an Euclidean distance space, using a triplet loss and a hard triplet mining strategy. It improves over previous work by employing an alternative input representation, and introducing a novel technique for temporal content summarization, a standardized latent space, and a data augmentation strategy specifically designed for VI. In addition to the main results, we perform an ablation study to highlight the importance of our design choices, and study the relation between embedding dimensionality and model performance.Index Terms-Cover song identification, deep learning, music embedding, network encoder.

show abstract

Triplet Convolutional Network for Music Version Identification

Cited by 8 publications

References 14 publications

Vocal-Accompaniment Compatibility Estimation Using Self-Supervised and Joint-Embedding Techniques

Vocal-Accompaniment Compatibility Estimation Using Self-Supervised and Joint-Embedding Techniques

Learning Low-Dimensional Embeddings of Audio Shingles for Cross-Version Retrieval of Classical Music

Accurate and Scalable Version Identification Using Musically-Motivated Embeddings

Contact Info

Product

Resources

About