Project Chimera
Fuente:
Zenodo
Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Recurso digital |
| Lingua: | inglese |
| Pubblicazione: |
Zenodo
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866901768191868928 |
|---|---|
| author | Olinger, Zack |
| author_facet | Olinger, Zack |
| contents | <h1> Project Chimera — Release Notes</h1> <h2>v1.0.0 — Initial Public Release</h2> <p> 2025-10-03</p> <p>This is the <strong>initial public release</strong> of <strong>Project Chimera</strong>, a modular audio transcription and analysis pipeline designed for long-form recordings, with support for flexible AI model integration via simple TOML configuration files.</p> <h3>✨ Features</h3> <ul> <li><p><strong>Two Core Scripts</strong></p> <ul> <li><code>transcribe_audio_by_date.py</code> — Handles date-based audio file naming and transcription using <a href="https://github.com/Purfview/whisper-standalone-win">Faster Whisper</a>.</li> <li><code>create_audio_journal_metadata.py</code> — Enriches transcription data with multi-model analytical layers, as defined in <code>analysis.toml</code>.</li> </ul> </li> <li><p><strong>Flexible Model Integration</strong></p> <ul> <li>AI models can be "dropped in" to the analysis pipeline by editing the <code>analysis.toml</code> configuration file.</li> <li>Example configuration and multiple model types are included to illustrate extensibility.</li> </ul> </li> <li><p><strong>Cross-Format Audio Support</strong></p> <ul> <li>Full MP3 support and partial FLAC support (with a single non-critical timestamp parsing error currently logged).</li> <li>Designed for long-form recordings, including sessions lasting many hours.</li> </ul> </li> <li><p><strong>Structured Output</strong></p> <ul> <li>Generates enriched JSON files containing transcriptions and analytical outputs.</li> <li>Optional CSV generation and metadata tagging for further downstream processing.</li> </ul> </li> <li><p><strong>Manual but Configurable Setup</strong></p> <ul> <li>Uses TOML-based configuration for transcription, analysis, and environment paths.</li> <li>Compatible with additional tooling such as Stanford CoreNLP and Swiss Ephemeris for extended analysis.</li> </ul> </li> <li><p><strong>Examples Included</strong></p> <ul> <li>The <code>example</code> folder provides sample outputs illustrating the structure of enriched JSON analysis files.</li> </ul> </li> </ul> <h3> Conceptual Foundation</h3> <p>Project Chimera is grounded in the <a href="https://github.com/whitelotusapps/Prolegomenon-of-Cybernetic-Shamanism"><strong>Prolegomenon of Cybernetic Shamanism</strong></a>, which provides the methodological framework that informed its design. Chimera functions as the analytical engine for transforming long-form spoken journals into structured insight data, feeding back into conceptual development.</p> <h3>⚠️ Known Limitations</h3> <ul> <li>Some paths and metadata are currently <strong>hard-coded</strong> for the developer's environment.</li> <li><strong>Manual folder structure creation</strong> is required prior to running the scripts.</li> <li>Analysis logging is currently <strong>incomplete</strong>.</li> <li>If all models are disabled in <code>analysis.toml</code>, the script raises an error rather than exiting gracefully.</li> <li>FLAC support works but emits a <strong>non-critical timestamp parsing error</strong>.</li> </ul> <h3> Licensing & Data Notes</h3> <ul> <li>Project Chimera is released under a <strong>Creative Commons Attribution–NonCommercial–ShareAlike 4.0 International License</strong>.</li> <li>The included <strong>XANEW dataset</strong> is covered by its original <strong>CC BY-NC-ND 3.0</strong> license. Usage of that dataset must comply with its terms.</li> </ul> <h2> Versioning</h2> <p>Project Chimera follows <strong>semantic versioning</strong>:</p> <p>MAJOR.MINOR.PATCH</p> <ul> <li><strong>MAJOR</strong> — Breaking changes to structure or configuration.</li> <li><strong>MINOR</strong> — New features, modules, or configuration options.</li> <li><strong>PATCH</strong> — Bug fixes and minor improvements.</li> </ul> <h2> Acknowledgments</h2> <ul> <li><a href="https://github.com/Purfview/whisper-standalone-win">Faster Whisper</a> for high-performance transcription on Windows.</li> <li><a href="https://stanfordnlp.github.io/CoreNLP/">Stanford CoreNLP</a> for NLP capabilities.</li> <li><a href="https://github.com/aloistr/swisseph">Swiss Ephemeris</a> for astrological calculations.</li> <li>Warriner, Kuperman, & Brysbaert (2013) for the <strong>XANEW dataset</strong>.</li> </ul> <p><em>© 2025 Zack Olinger — Project Chimera</em></p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_17260445 |
| institution | Zenodo |
| language | eng |
| publishDate | 2025 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | Project Chimera Olinger, Zack affective computing astrology audio analysis CoreNLP corpus analysis CSV data enrichment dataset generation emotional analysis emotional state emotional state analysis emotion analysis Faster Whisper feature extraction FLAC GPU acceleration idiolect idiolect analysis JSON language analysis large audio files lexical features linguistic analysis metadata extraction metaphyiscs model management modular model analysis MP3 multi model analysis natural language processing NLP Python PyTorch research tools Rich library self analysis self reflection shadow work signal processing Spacy speech corpus speech processing torchcrepe transcription analytics valence-arousal-dominance WSL2 XANEW zodiacal releasing <h1> Project Chimera — Release Notes</h1> <h2>v1.0.0 — Initial Public Release</h2> <p> 2025-10-03</p> <p>This is the <strong>initial public release</strong> of <strong>Project Chimera</strong>, a modular audio transcription and analysis pipeline designed for long-form recordings, with support for flexible AI model integration via simple TOML configuration files.</p> <h3>✨ Features</h3> <ul> <li><p><strong>Two Core Scripts</strong></p> <ul> <li><code>transcribe_audio_by_date.py</code> — Handles date-based audio file naming and transcription using <a href="https://github.com/Purfview/whisper-standalone-win">Faster Whisper</a>.</li> <li><code>create_audio_journal_metadata.py</code> — Enriches transcription data with multi-model analytical layers, as defined in <code>analysis.toml</code>.</li> </ul> </li> <li><p><strong>Flexible Model Integration</strong></p> <ul> <li>AI models can be "dropped in" to the analysis pipeline by editing the <code>analysis.toml</code> configuration file.</li> <li>Example configuration and multiple model types are included to illustrate extensibility.</li> </ul> </li> <li><p><strong>Cross-Format Audio Support</strong></p> <ul> <li>Full MP3 support and partial FLAC support (with a single non-critical timestamp parsing error currently logged).</li> <li>Designed for long-form recordings, including sessions lasting many hours.</li> </ul> </li> <li><p><strong>Structured Output</strong></p> <ul> <li>Generates enriched JSON files containing transcriptions and analytical outputs.</li> <li>Optional CSV generation and metadata tagging for further downstream processing.</li> </ul> </li> <li><p><strong>Manual but Configurable Setup</strong></p> <ul> <li>Uses TOML-based configuration for transcription, analysis, and environment paths.</li> <li>Compatible with additional tooling such as Stanford CoreNLP and Swiss Ephemeris for extended analysis.</li> </ul> </li> <li><p><strong>Examples Included</strong></p> <ul> <li>The <code>example</code> folder provides sample outputs illustrating the structure of enriched JSON analysis files.</li> </ul> </li> </ul> <h3> Conceptual Foundation</h3> <p>Project Chimera is grounded in the <a href="https://github.com/whitelotusapps/Prolegomenon-of-Cybernetic-Shamanism"><strong>Prolegomenon of Cybernetic Shamanism</strong></a>, which provides the methodological framework that informed its design. Chimera functions as the analytical engine for transforming long-form spoken journals into structured insight data, feeding back into conceptual development.</p> <h3>⚠️ Known Limitations</h3> <ul> <li>Some paths and metadata are currently <strong>hard-coded</strong> for the developer's environment.</li> <li><strong>Manual folder structure creation</strong> is required prior to running the scripts.</li> <li>Analysis logging is currently <strong>incomplete</strong>.</li> <li>If all models are disabled in <code>analysis.toml</code>, the script raises an error rather than exiting gracefully.</li> <li>FLAC support works but emits a <strong>non-critical timestamp parsing error</strong>.</li> </ul> <h3> Licensing & Data Notes</h3> <ul> <li>Project Chimera is released under a <strong>Creative Commons Attribution–NonCommercial–ShareAlike 4.0 International License</strong>.</li> <li>The included <strong>XANEW dataset</strong> is covered by its original <strong>CC BY-NC-ND 3.0</strong> license. Usage of that dataset must comply with its terms.</li> </ul> <h2> Versioning</h2> <p>Project Chimera follows <strong>semantic versioning</strong>:</p> <p>MAJOR.MINOR.PATCH</p> <ul> <li><strong>MAJOR</strong> — Breaking changes to structure or configuration.</li> <li><strong>MINOR</strong> — New features, modules, or configuration options.</li> <li><strong>PATCH</strong> — Bug fixes and minor improvements.</li> </ul> <h2> Acknowledgments</h2> <ul> <li><a href="https://github.com/Purfview/whisper-standalone-win">Faster Whisper</a> for high-performance transcription on Windows.</li> <li><a href="https://stanfordnlp.github.io/CoreNLP/">Stanford CoreNLP</a> for NLP capabilities.</li> <li><a href="https://github.com/aloistr/swisseph">Swiss Ephemeris</a> for astrological calculations.</li> <li>Warriner, Kuperman, & Brysbaert (2013) for the <strong>XANEW dataset</strong>.</li> </ul> <p><em>© 2025 Zack Olinger — Project Chimera</em></p> |
| title | Project Chimera |
| topic | affective computing astrology audio analysis CoreNLP corpus analysis CSV data enrichment dataset generation emotional analysis emotional state emotional state analysis emotion analysis Faster Whisper feature extraction FLAC GPU acceleration idiolect idiolect analysis JSON language analysis large audio files lexical features linguistic analysis metadata extraction metaphyiscs model management modular model analysis MP3 multi model analysis natural language processing NLP Python PyTorch research tools Rich library self analysis self reflection shadow work signal processing Spacy speech corpus speech processing torchcrepe transcription analytics valence-arousal-dominance WSL2 XANEW zodiacal releasing |
| url | https://doi.org/10.5281/zenodo.17260445 |