“Revisiting Meter Tracking in Carnatic Music using Deep Learning Approaches”
Satyajeet Prabhu · 2025
Read “Revisiting Meter Tracking in Carnatic Music using Deep Learning Approaches” by Satyajeet Prabhu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-11241.txt.
Introduction
Satyajeet Prabhu is the sole author of this 2025 music technology paper on meter tracking, the task of following the beat and the downbeat in a recording. Deep learning systems do this well for Western music, where labeled data are plentiful, but Carnatic music from South India, with its intricate rhythmic cycles called tālas, has barely been tested.
Prabhu evaluates two modern systems, a temporal convolutional network and a transformer model called Beat This!, against the probabilistic baseline used in earlier Carnatic work, on the standard Carnatic rhythm dataset. Off the shelf, the deep models did not always win, but after fine-tuning on Carnatic data and using musically informed settings, they matched or beat the baseline. The study shows mainstream models can be adapted to underrepresented traditions rather than rebuilt from scratch.
Catalog record
| CREATOR | Satyajeet Prabhu |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 80 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-11241
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
“Revisiting Meter Tracking in Carnatic Music using Deep Learning Approaches” Satyajeet Prabhu Abstract Abstract Beat and downbeat tracking, jointly referred to as Meter Tracking , is a fundamental task in Music Information Retrieval (MIR). Deep learning models have far surpassed traditional signal processing and classical machine learning approaches in this domain, particularly for Western (Eurogenetic) genres, where large annotated datasets are widely available. These systems, however, perform less reliably on underrepresented musical traditions. Carnatic music, a rich tradition from the Indian subcontinent, is renowned for its rhythmic intricacy and unique metrical structures (tāḷas) . The most notable prior work on meter tracking in this context employed probabilistic Dynamic Bayesian Networks (DBNs). The performance of state-of-the-art (SOTA) deep learning models on Carnatic music, however, remains largely unexplored. In this study, we evaluate two models for meter tracking in Carnatic music: the Temporal Convolutional Network (TCN), a lightweight architecture that has been successfully adapted for Latin rhythms, and Beat This! , a transformer-based model designed for broad stylistic coverage without the need for post-processing. Replicating the experimental setup of the DBN baseline on the Carnatic Music Rhythm (CMR f ) dataset, we systematically assess the performance of these models in a directly comparable setting. We further investigate adaptation strategies, including fine-tuning the models on Carnatic data and the use of musically informed parameters. Results show that while off-the-shelf models do not always outperform the DBN, their performance improves substantially with transfer learning, matching or surpassing the baseline. These findings indicate that SOTA deep learning models can be effectively adapted to underrepresented traditions, paving the way for more inclusive and broadly applicable meter tracking systems. Contents Chapter 1 Introduction 1.1 Background Rhythm analysis is a central topic in Music Information Retrieval (MIR), aimed at computationally analysing or modelling the temporal structure of music. It encompasses a variety of tasks such as onset detection, tempo estimation, beat and downbeat tracking, pattern analysis, microtiming analysis and synchronization among others, which together enable a comprehensive understanding of musical timing. This work focuses on the task of automatic estimation of beats and downbeats, commonly referred to as Meter Tracking , critical for several higher-level MIR tasks such as music segmentation and structural analysis, as well as applications such as DJ mixing and automatic beat matching. 1.1.1 Metrical Structure in Music Rhythm in music is perceived as pulsations organized at multiple hierarchical levels of differing timespans, known as its meter or metrical structure ( Bilmes 1993 ; London 2012 ) . These levels range from very fast subdivisions to larger organizational units (for example, see figure 1 ). The different metrical levels are described as follows: Figure 1 : Perceived metrical levels in ’ Twinkle, Twinkle, Little Star ’ Tatum The fastest regular pulse in the music that listeners can perceive as a meaningful subdivision of rhythm. Often corresponds to a 16th note in Western music, but the specific duration depends on the tempo and style. Tactus (Beat) The perceptually most salient pulse level that a listener would naturally tap their foot to. It typically corresponds to quarter notes in Western music but again depends on tempo and context. The beat level is central to most rhythm perception tasks and often serves as the reference level for tempo. Meter (Bar, Measure, Cycle) A grouping of beats into a recurring structure that establishes musical phrasing and form. Measures are typically marked by accent patterns and serve to shape listeners’ expectations of timing and emphasis. In Western music, meter is commonly represented using time signatures, such as 4/4, which indicates four beats per measure, with each beat typically being a quarter note in duration (see figure 2 ). Other common meters include 3/4 (e.g., waltz). Downbeat The first beat of a bar or cycle, often marked by a strong accent or structural change. It acts as a temporal anchor and plays a crucial role in conveying the start of a measure. Accurate downbeat perception is essential for understanding musical form and phrasing. Figure 2 : Musical meter in Western music. Figure from Wikipedia, Metre (music) . 1.1.2 Rhythm in Carnatic Music Carnatic music, one of the two principal traditions of Indian art music (IAM), is predominantly practiced and appreciated in the southern regions of the Indian subcontinent. It is distinguished by its dedicated audiences, sophisticated theoretical framework, and high level of musicianship. Traditionally, training in Carnatic music is transmitted orally through a lineage of teachers, with a strong emphasis on performance and improvisation. A typical Carnatic performance features a lead performer, a rhythmic accompaniment, a continuous background drone, and a melodic accompaniment. Unlike Western tonal music, Carnatic music does not employ harmony; instead, it is structured around the melodic framework of rāga and the rhythmic framework of tāḷa ( Sambamoorthy 1998 ) . Consequently, Carnatic music has become an important subject in MIR research as it presents unique challenges and opportunities for computational analysis ( Rao et al. 2023 ) . Rhythmic organization in Carnatic music is governed by the tāḷa system, a hierarchical framework of time cycles that underlies melodic and rhythmic phrasing as well as improvisation. Within each tāḷa cycle, sub-structures are defined to track progression through the cycle. While there are some conceptual parallels with Western metrical structures, the terminology and organization within the tāḷa system differ significantly. Table 1 presents approximate correspondences between metrical hierarchies in Western and Carnatic frameworks. Western Carnatic Tatum akṣara Beat (indicated by hand gestures) Measure āvartana Downbeat sama Table 1 : Mapping of Western and Carnatic rhythmic concepts Moreover, the tāḷa framework includes elements unique to the Carnatic tradition that lack direct equivalents in Western metrical theory, resulting in 175 theoretically possible tāḷas. In practice, however, a core set of 35 tāḷas is predominantly used in performance and pedagogy. Table 2 lists the four most commonly employed tāḷas in Carnatic music, alongside their total number of beats per cycle. Tāḷa #Beats Ādi 8 Rūpaka 3 Miśra chāpu 7 Khaṇḍa chāpu 5 Table 2 : Popular tāḷas in Carnatic music Figure 3 illustrates some of the concepts from table 1 using the example of an Ādi tāḷa cycle (8 beats). It also demonstrates how beats are further grouped into sections called aṅgas . The progression through the tāḷa cycle is marked by distinctive hand gestures, which indicate both individual beats and the different types of sections within the cycle. Figure 3 : Illustration of Ādi tāḷa. Figure from ( Srinivasamurthy 2016 ) The parallels outlined above, though not exact, provide a conceptual bridge between Western metrical hierarchies and Carnatic tāḷa organization. This, in turn, enables the adaptation of meter-tracking algorithms originally developed for Western music to the context of Carnatic music. 1.2 Motivation Over the past decade, deep learning approaches to meter tracking have significantly outperformed traditional signal processing and machine learning approaches. However, these models often excel primarily on Western (Eurogenetic) music genres, with limited generalisation to non-Western musical traditions due to several factors: 1. Data Representation and Bias : Most existing deep learning-based methods utilise supervised learning and require large amounts of high-quality annotated training data. Available datasets are often dominated by Western music. Creating such datasets for non-Western music is a challenging and expensive process. It often requires culturally