Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

Fretting-Transformer: Encoder-Decoder Model for MIDI to Tablature Transcription

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A compact T5 transformer converts MIDI into guitar tabs that agree with professional Guitar Pro tabs more often than A* or commercial tools on three datasets.

desk verdict A useful MIDI-to-tab system with a believable comparative result, but the tab-accuracy metric is too shaky to carry the central claim without more validation. read the letter →

arxiv 2506.14223 v1 pith:AESMVJB2 submitted 2025-06-17 cs.SD cs.CLcs.MMeess.AS

Anna Hamberger , Sebastian Murgul , Jochen Schmidt , Michael Heizmann This is my paper · ORCID
classification cs.SDcs.CLcs.MMeess.AS
keywords MIDItotablaturetranscriptionguitargenerationT5transformersymbolicmusictranslationstring-fretassignmentplayabilitymetriccapoandtuningconditioningProdatasets
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that guitar tablature transcription from MIDI can be solved as a text-to-text translation problem with a compact T5 encoder-decoder. The model takes MIDI note events as the source language and outputs combined string-fret tokens as the target language, with optional capo and tuning tokens as conditioning. On three test sets the paper reports tab accuracy of 98.41 percent on GuitarToday, 72.19 percent on Leduc, and 81.58 percent on DadaGP in standard tuning, beating the A* reimplementation, a lowest-fret baseline, Guitar Pro, and TuxGuitar in tab accuracy. The result matters because it suggests a learned model can replace hand-designed cost functions for producing playable guitar tabs, and because the same translation framing can absorb alternate tunings and capo positions without changing the architecture.

What carries the argument

The load-bearing object is the T5 text-to-text transformer, used as a reduced encoder-decoder (three layers, $d_{\text{model}}=128$, four attention heads) trained from scratch rather than initialized from a language model. The input language is an event-based MIDI stream with NOTE ON, NOTE OFF, and TIME SHIFT tokens; the output language is a sequence of TAB <string,fret> tokens, where each token encodes both choices at once. Capo and tuning are injected as CAPO and TUNING tokens in the conditioned variant. That arrangement turns ambiguous string-fret selection into a translation problem in which the decoder's attention can use a whole 20-note window of musical context, unlike the A* graph search, which only looks at adjacent notes. A post-processing step then maps model outputs back onto the original input pitches and selects viable string-fret pairs when no direct match exists.

What would settle it

Take a held-out set of professional tabs, ask experienced guitarists to blind-rank the model's tabs against A* and Guitar Pro, and check whether their preferences track the model's tab-accuracy margin; a complementary check is to retrain on DadaGP after excluding algorithmically generated tracks and see whether the Leduc margin survives.

Watch

Extended reading notes

Core claim

The paper's central claim is that a T5 encoder-decoder trained from scratch on a combined corpus of Guitar Pro files from DadaGP, GuitarToday, and Leduc produces tablatures that agree with the ground-truth Guitar Pro tabs more often than established methods do. In the standard-tuning scenario the Fretting-Transformer reaches 98.41% tab accuracy on GuitarToday, 72.19% on Leduc, and 81.58% on DadaGP, compared with 89.39% and 62.60% for A* on GuitarToday and Leduc; in the capo/tuning-conditioned scenario it again leads on GuitarToday (98.19%) and Leduc (73.02%), while GPT2 is slightly closer on DadaGP (81.45% vs. 79.89%). The paper also claims that a simple post-processing pass that matches predicted notes to input pitches restores 100% pitch accuracy, and that the best tokenization is the event-based v3 encoding with combined TAB tokens and TIME SHIFT tokens.

Load-bearing premise

The paper's evaluation assumes the ground-truth Guitar Pro tabs are the optimal playing positions, so the headline tab-accuracy numbers measure closeness to those tabs; if a substantial share of ground-truth tabs, especially in DadaGP, are not high-quality human arrangements, then 'surpassing the baselines' is really 'imitating one particular tab style' rather than producing objectively better tabs.

Editorial extensions

If this is right

  • If the central claim holds, a modest-size transformer without hand-coded biomechanical costs can outperform graph-search and commercial tab generators on tab accuracy across beginner, jazz, and multi-genre test sets.
  • Capo and tuning conditioning preserve most of the accuracy advantage, so the same model can produce tabs for common alternate tunings and capo positions without retraining per configuration.
  • The v3 encoding result implies that explicit timing information and a combined string-fret token are worth more than separate string and fret tokens for this task.
  • The post-processing result implies that a pure pitch-preservation guarantee is achievable at almost no cost to tab accuracy on the evaluated test sets, a prerequisite for any practical MIDI-to-tab tool.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper leaves implicit that tab accuracy against Guitar Pro ground truth is a measure of imitation of the corpus's fretting style; on DadaGP, where the paper notes tabs may be algorithmically generated, the Baseline's close scores suggest the metric may reward algorithmic patterns rather than human preference.
  • A blind listening or playing study with guitarists would separate imitation from playability, since the paper's own 'Smoke on the Water' example shows a divergence from ground truth that players might actually prefer.
  • The translation framing is generic: with a different target token vocabulary, the same encoder-decoder could be pointed at bass or ukulele tablature, or at lead-sheet-to-fingerstyle arrangement, because conditioning on tuning and capo is just extra tokens.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. This paper presents Fretting-Transformer, a T5-based encoder-decoder that maps symbolic MIDI note events to guitar tablature (string-fret tokens). The authors introduce preprocessing for three Guitar Pro datasets (GuitarToday, DadaGP, Leduc), compare five tokenization schemes, evaluate a capo/tuning-conditioned variant, and apply a post-processing correction step. The main empirical claim is that the proposed model achieves higher tablature accuracy than a baseline, A*, Guitar Pro, and TuxGuitar on all three test sets, with additional difficulty scores reported. The manuscript is clearly written and presents a practical system, but the central evaluation is weakened by a circular pitch-accuracy metric and by reliance on a tab-accuracy ground truth whose optimality the authors themselves question.

Significance. The paper addresses a genuine and practical problem in MIR: converting MIDI to guitar tablature requires resolving string-fret ambiguity while preserving playability. The work is useful as an engineering contribution: it demonstrates that a compact T5 model trained from scratch can produce plausible tabs, and it provides a systematic comparison of input/output encodings and conditioning strategies. It also includes an honest discussion of dataset-quality limitations, and the DadaGP filtering and augmentation pipeline may be reusable. However, the headline claim of surpassing all baselines is not yet established: the pitch-accuracy result is guaranteed by construction, tab accuracy is measured against ground truth of questionable optimality, and no statistical significance is reported. A human-evaluation study or a metric validation against guitarist preference would be needed to make the central claim convincing.

major comments (4)
  1. [Section 3.5 / Table 2] Section 3.5 defines post-processing as matching each input note to the closest output note within a window of ±5 notes and, if no direct match is found, applying 'the first viable string-fret combination generated for the guitar configuration used.' Consequently, the 100.00% pitch accuracy in Table 2 is not a property of the learned model but a guarantee of the correction procedure; reporting it as an evaluation result is circular. Please report the raw model pitch accuracy before post-processing, and, if pitch fidelity is presented as a system property, state it as a design guarantee rather than as measured model quality.
  2. [Section 3.6 / Table 5] The central comparison rests on tab accuracy, which counts exact agreement with ground-truth string-fret pairs that are 'assumed to represent the optimal playing positions.' The manuscript itself provides reasons to doubt this assumption: Section 3.1 says DadaGP contributions 'vary greatly,' Section 4.6 says 'a lot of the tablatures in this dataset are algorithmically generated,' and Figure 3 (box 1) shows a Fretting-Transformer output that deviates from the GuitarToday ground truth and 'might be preferred by many guitarists.' A metric that penalizes such arguably better fingerings cannot by itself support the claim that the model surpasses baselines in transcription quality. At minimum, provide an analysis of agreement with human preference or playability judgments, and separate results on professionally curated subsets from the crowd-sourced and algorithmically generated subsets.
  3. [Table 5] All reported accuracies are point estimates from a single evaluation run, with no confidence intervals, standard deviations, or significance tests. This matters particularly for the small datasets: the Leduc test split is roughly 23 files (10% of 232), and on GuitarToday the gap between Ours (98.41%) and Baseline/TuxGuitar (98.30%) is 0.11 percentage points. Please add multiple-seed runs, paired tests based on per-file differences, and report variance or confidence intervals; otherwise phrases like 'significant improvements' in Section 4.6 are not supported by the evidence presented.
  4. [Section 4.6 / Table 5] The paper reports higher difficulty for the proposed model than for A* on Leduc (4.2415 vs. 2.5704) while still claiming the model is superior, and it dismisses the trade-off by saying 'an exclusive focus on playability may not reflect the preferences of guitarists.' This is an assertion, not evidence. If the claim is that higher tab accuracy is worth a harder tab, the authors need a preference study or an argument about why the difficulty metric mis-orders quality in these cases. Without that, the difficulty results undercut rather than support the playability claim.
minor comments (5)
  1. [Section 3.4] The text says the model halves the configuration of t5-small but gives d_model=128, d_ff=1024, three layers, and four attention heads; since t5-small has d_model=512, d_ff=2048, six layers, and eight heads, 'halving' is imprecise (the dimensions are quartered). Please clarify the actual configuration.
  2. [Section 3.5 / Section 4.2] Section 4.2 refers to two post-processing variants, 'overlap correction' and 'neighbor search,' but Section 3.5 describes only one matching algorithm; please define the two variants explicitly.
  3. [Section 3.6] In Equations (4)–(6), the same symbols p and q are used for fret numbers in some equations and string numbers in others, which makes the definitions hard to follow; please use distinct notation for frets and strings.
  4. [Figure 2] The y-axis label reads 'T ab Accuracy' with a missing space; please fix the typo.
  5. [Section 4.4] The heading 'Domain Adaption from Text' should read 'Domain Adaptation from Text.'

Circularity Check

1 steps flagged · score 4.0 of 10

The 100% pitch-accuracy result is self-fulfilling because post-processing forces notes to match the input pitches, but the central tab-accuracy comparison against A*, Guitar Pro, and TuxGuitar uses external ground truth and is not circular.

  1. self definitional [Section 3.5 Data Post-Processing; Section 4.2 / Table 2]
    "Our post-processing algorithm refines the model's output by comparing the estimated note sequence to the corresponding input note sequence. It attempts to match each input note to its closest counterpart in the estimated sequence within a configurable window of ±5 notes. ... If no direct match is found, the first viable string-fret combination generated for the guitar configuration used is applied. That way, we ensure that the tablatures reflect the original notes."

    Pitch accuracy is defined in Section 3.6 as how well the model reproduces the original pitches from the MIDI input. The post-processing step explicitly forces each output note to match an input note, substituting a pitch-correct string/fret combination when needed. Hence the 100.00% pitch accuracy reported in Table 2 after 'Overlap + Neighbor Search' is guaranteed by the algorithm rather than being an independent property of the model. The metric cannot fail after this post-processing; it is a self-fulfilling result. Tab accuracy, in contrast, is measured against held-out ground-truth tabs and is not adjusted to match the ground truth, so the central comparison remains non-circular.

full rationale

The only concrete circularity is the pitch-accuracy metric after post-processing. Section 3.5 states that the post-processor matches every estimated note to the corresponding input note and, if necessary, substitutes the first valid string-fret combination, 'so that the tablatures reflect the original notes'; Section 3.6 defines pitch accuracy as reproducing the original MIDI pitches. Therefore the 100.00% pitch accuracy in Table 2 is a guaranteed consequence of the algorithm, not a measured prediction. This does not extend to the paper's main comparative claim: tab accuracy is computed against held-out Guitar Pro ground-truth tabs, and neither post-processing nor training was conditioned on those test labels, so the Fretting-Transformer's advantage over A*, Guitar Pro, and TuxGuitar in Table 5 is an externally benchmarked result. The paper's own caveat that ground truth is only 'assumed to represent the optimal playing positions' and its Figure 3 example where open strings might be preferable are validity concerns about the metric, not circular reductions. No load-bearing self-citations or imported uniqueness theorems appear; the transformer architecture is standard T5 trained from scratch. Score 4 reflects one self-fulfilling secondary metric alongside an otherwise independent central benchmark.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

All load-bearing assumptions are domain or data-quality assumptions rather than new mathematical axioms. The model introduces no new physical entities. The main risk is that the metrics and the ground truth are trusted as objective quality measures despite acknowledged imperfections in the datasets and a heuristic playability score.

free parameters (5)
  • Post-processing search window = ±5 notes
    The overlap and neighbor search in Section 3.5 uses a configurable window to match model outputs to input notes. Hand-chosen; changing it changes tab accuracy and the final comparison.
  • Difficulty metric weights = 0.25, 0.50, 0.75 (from Radicioni-Lombardo)
    The playability score in Section 3.6 uses fixed costs for fret and string movement rather than learned or measured values.
  • Locality factor alpha = 0.25
    Set by tests in the original difficulty framework [25] and reused without refitting in Equation 5.
  • Reduced T5 hyperparameters = d_model=128, d_ff=1024, 3 layers, 4 heads
    Model size and depth are chosen by the authors; performance and convergence depend on them (Section 3.4).
  • Augmentation ranges = capo 0-7; tunings standard, half-step, full-step, drop-D
    Data augmentation choices in Section 3.3 are hand-set and determine the conditioned model's coverage.
assumptions (5)
  • domain assumption Ground-truth Guitar Pro tablatures represent optimal or preferred playing positions
    Invoked in Section 3.6 when defining tab accuracy, and acknowledged as not always true in Section 4.7.
  • domain assumption Guitar Pro to MIDI conversion and channel-based filtering preserve pitch, timing, string, and fret information
    Section 3.2 relies on MIDI channel IDs 25/26 and keyword filters to select acoustic guitar tracks; errors in conversion or filtering would propagate to training data.
  • domain assumption The difficulty metric of Radicioni-Lombardo is a valid proxy for human playability
    Section 3.6 uses the metric to compare models without a human listening or playing test.
  • domain assumption The post-processor can always find a valid string-fret combination for any input pitch under the target tuning and capo
    Section 3.5 assumes a first viable string-fret combination exists for every note; this is true only for pitches within the guitar's range and configuration.
  • domain assumption T5 text-to-text framing is suitable for symbolic music translation
    Section 3.4 adopts the T5 architecture without formal justification beyond prior NLP success.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Fretting-Transformer: Encoder-Decoder Model for MIDI to Tablature Transcription." pith.science (2026). https://pith.science/paper/AESMVJB2

@misc{pith2026250614223,
  author       = {Pith},
  title        = {Pith review of: Fretting-Transformer: Encoder-Decoder Model for MIDI to Tablature Transcription},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AESMVJB2}},
  note         = {Machine review of arXiv:2506.14223}
}
read the original abstract

Music transcription plays a pivotal role in Music Information Retrieval (MIR), particularly for stringed instruments like the guitar, where symbolic music notations such as MIDI lack crucial playability information. This contribution introduces the Fretting-Transformer, an encoderdecoder model that utilizes a T5 transformer architecture to automate the transcription of MIDI sequences into guitar tablature. By framing the task as a symbolic translation problem, the model addresses key challenges, including string-fret ambiguity and physical playability. The proposed system leverages diverse datasets, including DadaGP, GuitarToday, and Leduc, with novel data pre-processing and tokenization strategies. We have developed metrics for tablature accuracy and playability to quantitatively evaluate the performance. The experimental results demonstrate that the Fretting-Transformer surpasses baseline methods like A* and commercial applications like Guitar Pro. The integration of context-sensitive processing and tuning/capo conditioning further enhances the model's performance, laying a robust foundation for future developments in automated guitar transcription.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Explicit Note-Event Tokenization and Pitch-Validity Constrained Decoding for MIDI-to-Tablature Transcription

    cs.SD 2026-07 conditional novelty 6.0 of 10

    Adding explicit note-event tokens to the decoder and masking pitch-invalid TAB positions during decoding improves guitar tablature accuracy over the Fretting Transformer on DadaGP and especially on the small François ...

Reference graph

Works this paper leans on

32 extracted references · 29 canonical work pages · cited by 1 Pith paper

  1. [1]

    This task is both challenging and essential in the field of Music Information Retrieval (MIR) [1]

    INTRODUCTION Notation-level music transcription is the process of con- verting musical audio or symbolic data into a written form. This task is both challenging and essential in the field of Music Information Retrieval (MIR) [1]. Automatic Music Transcription (AMT) seeks to address the limitations of manual transcription by creating algorithms that can tr...

    work page 2025

  2. [2]

    Fretting-Transformer: Encoder-Decoder Model for MIDI to Tablature Transcription

    RELATED WORK There are various approaches in the field of automatic guitar tablature transcription, including rule-based, probabilistic, graph-based, and neural network-based methods. Early sys- tems relied on predefined rules to create tablatures from MIDI data. The software developed by Wang and Li uti- lizes harmonic rules and fretting styles to produc...

    work page Pith review arXiv 2025

  3. [3]

    This format allows for direct experimentation with MIDI tabla- ture transcription by encoding pitch as well as string/fret information

    METHODOLOGY 3.1 Datasets Unlike datasets that concentrate on converting audio into symbolic formats through complex Automatic Music Tran- scription (AMT) pipelines [ 1], the datasets used in this research – DadaGP, GuitarToday and Leduc – focus on symbolic data provided in the Guitar Pro 1 format. This format allows for direct experimentation with MIDI ta...

    work page 2020

  4. [4]

    For evaluation, the metrics described in Section 3.6 are used

    EXPERIMENTS AND RESULTS In this section, the results of our proposed model are pre- sented in an evaluation on the test split of the GuitarToday, Leduc and DadaGP datasets. For evaluation, the metrics described in Section 3.6 are used. 4.1 Data Encodings To evaluate the effects of data encoding on the transcription performance of guitar tablatures, we con...

    work page

  5. [5]

    By framing the task as a symbolic translation problem, the Fretting-Transformer outperforms existing methods, including A∗ and commer- cial tools like Guitar Pro

    CONCLUSIONS This research demonstrates the effectiveness of transformer models, particularly the T5 architecture, in transcribing MIDI sequences into guitar tablatures. By framing the task as a symbolic translation problem, the Fretting-Transformer outperforms existing methods, including A∗ and commer- cial tools like Guitar Pro. Key contributions include...

    work page

  6. [6]

    Comparative Study of Genetic Algorithm and Ant Colony Optimization Algorithm Performances for the Task of Guitar Tablature Transcription,

    J. V . Ramos, A. S. Ramos, C. N. Silla, and D. S. Sanches, “Comparative Study of Genetic Algorithm and Ant Colony Optimization Algorithm Performances for the Task of Guitar Tablature Transcription,” in2015 Brazilian Conference on Intelligent Systems (BRACIS). IEEE, 2015, pp. 228–233

    work page 2015

  7. [7]

    Auto- matic Music Transcription: An Overview,

    E. Benetos, S. Dixon, Z. Duan, and S. Ewert, “Auto- matic Music Transcription: An Overview,”IEEE Signal Processing Magazine, vol. 36, no. 1, pp. 20–30, 2018

    2018

  8. [8]

    Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer,

    C. Raffel, N. Shazeer, A. Roberts, K. Lee, S. Narang, M. Matena, Y . Zhou, W. Li, and P. J. Liu, “Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer,”Journal of Machine Learning Research, vol. 21, no. 140, pp. 1–67, 2020

    2020

Show all 32 references
  1. [9]

    Generating Guitar Scores from a MIDI Source,

    J.-F. Wang and T.-Y . Li, “Generating Guitar Scores from a MIDI Source,” inProceedings of 1997 Interna- tional Symposium on Multimedia Information Process- ing, Taipei, Taiwan, 1997

    1997

  2. [10]

    Con- structing a system for finger–position determination and tablature generation for playing melodies on guitars,

    M. Miura, I. Hirota, N. Hama, and M. Yanagida, “Con- structing a system for finger–position determination and tablature generation for playing melodies on guitars,” Systems and Computers in Japan, vol. 35, no. 6, pp. 10–19, 2004

    2004

  3. [11]

    A genetic algorithm for the automatic generation of playable guitar tablature,

    D. R. Tuohy and W. D. Potter, “A genetic algorithm for the automatic generation of playable guitar tablature,” inProceedings of the International Computer Music Conference, 2005, pp. 499–502

    2005

  4. [12]

    Guitar Tablature Estimation with a Convolutional Neural Network,

    A. Wiggins and Y . Kim, “Guitar Tablature Estimation with a Convolutional Neural Network,” inProceedings of the 20th International Society for Music Information Retrieval Conference, 2019, pp. 284–291

    2019

  5. [13]

    A Few-Sample Strategy for Guitar Tablature Transcription Based on Inharmonic- ity Analysis and Playability Constraints,

    G. Bastas, S. Koutoupis, M. Kaliakatsos-Papakostas, V . Katsouros, and P. Maragos, “A Few-Sample Strategy for Guitar Tablature Transcription Based on Inharmonic- ity Analysis and Playability Constraints,” inICASSP 2022-2022 IEEE International Conference on Acoustics, Speech an...

    2022

  6. [14]

    Automatic Transcription of Guitar Chords and Fingering From Audio,

    A. M. Barbancho, A. Klapuri, L. J. Tardon, and I. Bar- bancho, “Automatic Transcription of Guitar Chords and Fingering From Audio,”IEEE Transactions on Audio, Speech, and Language Processing, vol. 20, no. 3, pp. 915–921, 2012

    2012

  7. [15]

    Fingering for String Instruments with the Optimum Path Paradigm,

    S. I. Sayegh, “Fingering for String Instruments with the Optimum Path Paradigm,”Computer Music Journal, vol. 13, no. 3, p. 76, 1989

    1989

  8. [16]

    A segmentation-based prototype to compute string instru- ments fingering,

    D. Radicioni, L. Anselma, and V . Lombardo, “A segmentation-based prototype to compute string instru- ments fingering,” inProceedings of the Conference on Interdisciplinary Musicology, vol. 17, 2004, p. 97

    2004

  9. [17]

    Robotaba Guitar Tablature Transcription Framework,

    G. Burlet and I. Fujinaga, “Robotaba Guitar Tablature Transcription Framework,” inProceedings of the 14th International Society for Music Information Retrieval Conference, 2013, pp. 517–522

    2013

  10. [18]

    Automatic composition of guitar tabs by transformers and groove modeling,

    Y .-H. Chen, Y .-H. Huang, W.-Y . Hsiao, and Y .-H. Yang, “Automatic composition of guitar tabs by transformers and groove modeling,”Proceedings of the 21th Interna- tional Society for Music Information Retrieval Confer- ence, pp. 756–763, 2020

    2020

  11. [19]

    Note-level Automatic Guitar Transcription Using Attention Mechanism,

    S. Kim, T. Hayashi, and T. Toda, “Note-level Automatic Guitar Transcription Using Attention Mechanism,” in 2022 30th European Signal Processing Conference (EU- SIPCO), 2022, pp. 229–233

    2022

  12. [20]

    Music Transformer: Gen- erating Music with Long-Term Structure,

    C.-Z. A. Huang, A. Vaswani, J. Uszkoreit, N. Shazeer, I. Simon, C. Hawthorne, A. M. Dai, M. D. Hoffman, M. Dinculescu, and D. Eck, “Music Transformer: Gen- erating Music with Long-Term Structure,” inInterna- tional Conference on Learning Representations, 2018

    2018

  13. [21]

    Pop Music Transformer: Beat-based Modeling and Generation of Expressive Pop Piano Compositions,

    Y .-S. Huang and Y .-H. Yang, “Pop Music Transformer: Beat-based Modeling and Generation of Expressive Pop Piano Compositions,” inProceedings of the 28th ACM International Conference on Multimedia. New York, NY , USA: Association for Computing Machinery, 2020, pp. 1180–1188

    2020

  14. [22]

    Theme Transformer: Symbolic Music Gen- eration With Theme-Conditioned Transformer,

    Y .-J. Shih, S.-L. Wu, F. Zalkow, M. M ¨uller, and Y .- H. Yang, “Theme Transformer: Symbolic Music Gen- eration With Theme-Conditioned Transformer,”IEEE Transactions on Multimedia, vol. 25, pp. 3495–3508, 2022

    2022

  15. [23]

    DadaGP: A Dataset of Tokenized GuitarPro Songs for Sequence Models,

    P. Sarmento, A. Kumar, C. J. Carr, Z. Zukowski, M. Bar- thet, and Y .-H. Yang, “DadaGP: A Dataset of Tokenized GuitarPro Songs for Sequence Models,” inProceedings of the 22th International Society for Music Information Retrieval Conference, 2021, pp. 610–617

    2021

  16. [24]

    Franc ¸ois Leduc Online Library,

    “Franc ¸ois Leduc Online Library,” https: //www.francoisleduconlinelibrary.com/, accessed: 2024-11-14

    2024

  17. [25]

    The scoring system takes into account two types of movement: horizontal shifts along the fretboard (along) and vertical shifts across the strings (across)

    is used to objectively evaluate the playability of tab- latures. The scoring system takes into account two types of movement: horizontal shifts along the fretboard (along) and vertical shifts across the strings (across). The difficulty of transitioning between two positions (p...

  18. [26]

    Transformer-XL: Attentive Lan- guage Models Beyond a Fixed-Length Context,

    Z. Dai, Z. Yang, Y . Yang, J. Carbonell, Q. V . Le, and R. Salakhutdinov, “Transformer-XL: Attentive Lan- guage Models Beyond a Fixed-Length Context,” in Proceedings of the 57th Annual Meeting of the Associ- ation for Computational Linguistics. Florence, Italy: Association for...

    2019

  19. [27]

    MIDI-to-Tab: Guitar Tablature Inference via Masked Language Modeling,

    D. Edwards, X. Riley, P. Sarmento, and S. Dixon, “MIDI-to-Tab: Guitar Tablature Inference via Masked Language Modeling,” https://arxiv.org/abs/2408.05024, 2024, accessed: 2024-11-14

    2024 arXiv

  20. [28]

    BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding,

    J. Devlin, M. Chang, K. Lee, and K. Toutanova, “BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding,” inProceedings of the 2019 Conference of the North American Chapter of the As- sociation for Computational Linguistics: Human Lan- guage Technologi...

    2019

  21. [29]

    miditok: A Python package for MIDI file tokenization,

    N. Fradet, J.-P. Briot, F. Chhel, A. E. F. Seghrouchni, and N. Gutowski, “miditok: A Python package for MIDI file tokenization,” http://arxiv.org/abs/2310. 17202, accessed: 2024-11-14

    2024

  22. [30]

    High resolution guitar transcription via domain adaptation,

    X. Riley, D. Edwards, and S. Dixon, “High resolution guitar transcription via domain adaptation,” inICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2024, pp. 1051–1055

    2024

  23. [31]

    Guitar Fingering for Music Performance,

    D. P. Radicioni and V . Lombardo, “Guitar Fingering for Music Performance,” inInternational Conference on Mathematics and Computing, 2005

    2005

  24. [32]

    Language Models are Unsupervised Multitask Learners,

    A. Radford, J. Wu, R. Child, D. Luan, D. Amodei, and I. Sutskever, “Language Models are Unsupervised Multitask Learners,” 2019

    2019

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.