From 71cf3b6fb0ffaba881995f1700bc9b41d3d352fd Mon Sep 17 00:00:00 2001 From: genisplaja Date: Fri, 29 Nov 2024 17:55:26 +0100 Subject: [PATCH] fixes in build, increase timeout --- webbook/_config.yml | 17 +++++++++++++++-- webbook/bibliography/separation.bib | 2 +- webbook/resources/exploring-performance.ipynb | 10 +++++----- .../exploring-raga.ipynb | 0 webbook/rhythmic_analysis/meter_analysis.ipynb | 1 + .../structure_analysis/music-segmentation.ipynb | 2 +- 6 files changed, 23 insertions(+), 9 deletions(-) rename webbook/{melodic_analysis => resources}/exploring-raga.ipynb (100%) diff --git a/webbook/_config.yml b/webbook/_config.yml index 53fdf5c..577a448 100644 --- a/webbook/_config.yml +++ b/webbook/_config.yml @@ -10,7 +10,7 @@ logo: images/logo.png execute: execute_notebooks: force allow_errors: true - timeout: 5000 + timeout: 10000 # Define the name of the latex output file for PDF builds latex: @@ -32,6 +32,7 @@ bibtex_bibfiles: - bibliography/rhythmic_analysis.bib - bibliography/structure_analysis.bib - bibliography/timbre_analysis.bib + - bibliography/separation.bib # Information about where the book exists on the web @@ -47,4 +48,16 @@ html: use_repository_button: true # Exclude this patterns from build -exclude_patterns: [.github/*, .venv/*, webbook/carnatic/*, webbook/hindustani/*, webbook/separation/misc/*, webbook/resources/exploring-raga.ipynb] \ No newline at end of file +exclude_patterns: + - .github/* + - .venv/* + - webbook/carnatic/* + - webbook/hindustani/* + - webbook/separation/misc/* + - webbook/resources/exploring-raga.ipynb + - webbook/separation/misc/carnatic-separation-leakage-finetune.ipynb + - webbook/separation/misc/carnatic-separation-leakage.ipynb + - webbook/separation/misc/carnatic-separation-spleeter.ipynb + - webbook/separation/misc/intro-carnatic-music.md + - webbook/resources/our-timeline.md + - README.md \ No newline at end of file diff --git a/webbook/bibliography/separation.bib b/webbook/bibliography/separation.bib index 44e752c..4492453 100644 --- a/webbook/bibliography/separation.bib +++ b/webbook/bibliography/separation.bib @@ -10,7 +10,7 @@ @article{spleeter title = {Spleeter: a fast and efficient music source separation tool with pre-trained models}, journal = {Journal of Open Source Software}, doi = {10.21105/joss.02154}} -@article{demucs +@article{demucs, author = {Défossez, Alexandre and Usunier, Nicolas and Bottou, Léon and Bach, Francis}, eprint = {1911.13254}, title = {Music Source Separation in the Waveform Domain}, diff --git a/webbook/resources/exploring-performance.ipynb b/webbook/resources/exploring-performance.ipynb index 952c13f..5b6146b 100644 --- a/webbook/resources/exploring-performance.ipynb +++ b/webbook/resources/exploring-performance.ipynb @@ -25,7 +25,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "## Import Dependencies and Data" + "## 1. Import Dependencies and Data" ] }, { @@ -291,7 +291,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "# 1. Loading and visualising the data" + "## 2. Loading and visualising the data" ] }, { @@ -577,7 +577,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "# 2. Feature Extraction" + "## 2. Feature Extraction" ] }, { @@ -1287,7 +1287,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "## Percussion onset detection" + "## 3. Rhythm analysis: Percussion onset detection" ] }, { @@ -1407,7 +1407,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "### 2.6 Melodic pattern discovery" + "## 4. Melodic analysis: Melodic pattern discovery" ] }, { diff --git a/webbook/melodic_analysis/exploring-raga.ipynb b/webbook/resources/exploring-raga.ipynb similarity index 100% rename from webbook/melodic_analysis/exploring-raga.ipynb rename to webbook/resources/exploring-raga.ipynb diff --git a/webbook/rhythmic_analysis/meter_analysis.ipynb b/webbook/rhythmic_analysis/meter_analysis.ipynb index c277fa2..3979ad5 100644 --- a/webbook/rhythmic_analysis/meter_analysis.ipynb +++ b/webbook/rhythmic_analysis/meter_analysis.ipynb @@ -6,6 +6,7 @@ "source": [ "(meter-analysis)=\n", "# Meter analysis\n", + "\n", "In the musical introduction, the main rhythmic characteristics of Carnatic and Hindustani traditions are provided. The presented hierarchical form of a tala is targeted in {cite}`srinivasamurthy_rhythm_2014`, using a supervised approach on top of spectral hand-crafted features that capture tempo and onsets in the signal. Bayesian models are found to be the most capable of capture the rhythmic cycles in music signals, first applied in the form of particle filters {cite}`srinivasamurthy_rhythm_2015` for an efficient inference, showing competetitive results on Carnatic recordings. The said approach is futher extended to a generalized model to capture long cycles {cite}`srinivasamurthy_rhythm_2016`, being this very convenient for rhythm tracking in Hindustani Music." ] }, diff --git a/webbook/structure_analysis/music-segmentation.ipynb b/webbook/structure_analysis/music-segmentation.ipynb index 7adfba4..229174f 100644 --- a/webbook/structure_analysis/music-segmentation.ipynb +++ b/webbook/structure_analysis/music-segmentation.ipynb @@ -403,7 +403,7 @@ "## A note on source separation for Indian Art Music\n", "As you may have noticed in the displayed audio above, even though `spleeter` is within the best out-of-the-box source separation models to use, we need to take into account some considerations in regards to the task of music source separation for Indian Art Music signals.\n", "\n", - "First of all, although `spleeter` is trained with a massive amount of recordings, we can safely assume that Carnatic and Hindustani music do not have a considerable representation in the training set (this applies to the other out-of-the-box source separation models out there). In that sense, it is expected that these models struggle to generalize to Indian Art Music specific instruments and arrangements, which may cause abnormally low interference removal performance. The predominance of melodic monophonic accompaniment instruments in both [Carnatic](carnatic-instrumentation) (the violin being the most common case) and [Hindustani](hindustani-instrumentation) (harmonium in this case), the [tambura drone](carnatic-tambura-drone), the pitched percussion... These are high-level examples of elements that may cause the standardized source separation models to not generalize properly to Indian Art Music signals.\n", + "First of all, although `spleeter` is trained with a massive amount of recordings, we can safely assume that Carnatic and Hindustani music do not have a considerable representation in the training set (this applies to the other out-of-the-box source separation models out there). In that sense, it is expected that these models struggle to generalize to Indian Art Music specific instruments and arrangements, which may cause abnormally low interference removal performance. The predominance of melodic monophonic accompaniment instruments in both [Carnatic](carnatic-instrumentation) (the violin being the most common case) and [Hindustani](hindustani-repertoire) (harmonium in this case), the [tambura drone](carnatic-tambura-drone), the pitched percussion... These are high-level examples of elements that may cause the standardized source separation models to not generalize properly to Indian Art Music signals.\n", "\n", "What is more, the standardized source separation models target whether *vocals* and *accompaniment*, or *vocals*, *bass*, *drums*, and *other*. While to separate the singing voice from an accompaniment is OK, the *4 stem* configuration is far from being representative of the actual Carnatic and Hindustani Music arrangements.\n", "\n",