{"data":{"slug":"cstr-edinburgh-merlin","name":"merlin","tagline":"Neural Network based Speech Synthesis System","github_url":"https://github.com/CSTR-Edinburgh/merlin","owner":"CSTR-Edinburgh","repo":"merlin","owner_avatar_url":"https://avatars.githubusercontent.com/u/14782539?v=4","primary_language":"Python","stars":1319,"forks":431,"topics":["deep-learning","keras","merlin","python","speech-synthesis","tensorflow","text-to-speech","theano","voice-conversion"],"archived":false,"github_pushed_at":"2020-03-03T10:46:34+00:00","maintenance_label":"Dormant","url":"https://www.graphcanon.com/tools/cstr-edinburgh-merlin","markdown_url":"https://www.graphcanon.com/tools/cstr-edinburgh-merlin.md","api_url":"https://www.graphcanon.com/api/graphcanon/tools/cstr-edinburgh-merlin","graph_url":"https://www.graphcanon.com/api/graphcanon/graph?tool=cstr-edinburgh-merlin","description":"This is now the official location of the Merlin project.","homepage_url":"http://www.cstr.ed.ac.uk/projects/merlin/","license":"Apache-2.0","open_issues":207,"watchers":99,"ai_summary":"Merlin is a toolkit for building Deep Neural Network models for statistical parametric speech synthesis. It uses Python and relies on Theano, with optional support for TensorFlow and Keras.","readme_excerpt":"## Merlin: The Neural Network (NN) based Speech Synthesis System\n\nThis repository contains the Neural Network (NN) based Speech Synthesis System  \ndeveloped at the Centre for Speech Technology Research (CSTR), University of \nEdinburgh. \n\nMerlin is a toolkit for building Deep Neural Network models for statistical parametric speech synthesis. \nIt must be used in combination with a front-end text processor (e.g., Festival) and a vocoder (e.g., STRAIGHT or WORLD).\n\nThe system is written in Python and relies on the Theano numerical computation library.\n\nMerlin comes with recipes (in the spirit of the [Kaldi](https://github.com/kaldi-asr/kaldi) automatic speech recognition toolkit) to show you how to build state-of-the art systems.\n\nMerlin is free software, distributed under an Apache License Version 2.0, allowing unrestricted commercial and non-commercial use alike.\n\nRead the documentation at [cstr-edinburgh.github.io/merlin](https://cstr-edinburgh.github.io/merlin/).\n\nMerlin is compatible with: __Python 2.7-3.6__.\n\nInstallation\n------------\n\nMerlin uses the following dependencies:\n\n- numpy, scipy\n- matplotlib\n- bandmat\n- theano\n- tensorflow (optional, required if you use tensorflow models)\n- sklearn, keras, h5py (optional, required if you use keras models)\n\nTo install Merlin, `cd` merlin and run the below steps:\n\n- Install some basic tools in Merlin\n```sh\nbash tools/compile_tools.sh\n```\n- Install python dependencies\n```sh\npip install -r requirements.txt\n```\n\nFor detailed instructions, to build the toolkit: see [INSTALL](https://github.com/CSTR-Edinburgh/merlin/blob/master/INSTALL.md) and [CSTR blog post](https://cstr-edinburgh.github.io/install-merlin/).  \nThese instructions are valid for UNIX systems including various flavors of Linux;\n\n\nGetting started with Merlin\n---------------------------\n\nTo run the example system builds, see `egs/README.txt`\n\nAs a first demo, please follow the scripts in `egs/slt_arctic`\n\nNow, you can also follow Josh Meyer's [blog post](http://jrmeyer.github.io/tts/2017/02/14/Installing-Merlin.html) for detailed instructions <br/> on how to install Merlin and build SLT demo voice.\n\nFor a more in-depth tutorial about building voices with Merlin, you can check out:\n\n- [Deep Learning for Text-to-Speech Synthesis, using the Merlin toolkit (Interspeech 2017 tutorial)](http://www.speech.zone/courses/one-off/merlin-interspeech2017)\n- [Arctic voices](https://cstr-edinburgh.github.io/merlin/getting-started/slt-arctic-voice)\n- [Build your own voice](https://cstr-edinburgh.github.io/merlin/getting-started/build-own-voice)\n\n\nSynthetic speech samples\n------------------------\n\nListen to [synthetic speech samples](https://cstr-edinburgh.github.io/merlin/demo.html) from our SLT arctic voice.\n\nDevelopment pattern for contributors\n------------------------------------\n\n1. [Create a personal fork](https://help.github.com/articles/fork-a-repo/)\nof the [main Merlin repository](https://github.com/CSTR-Edinburgh/merlin) in GitHub.\n2. Make your changes in a named branch different from `master`, e.g. you create\na branch `my-new-feature`.\n3. [Generate a pull request](https://help.github.com/articles/creating-a-pull-request/)\nthrough the Web interface of GitHub.\n\nContact Us\n----------\n\nPost your questions, suggestions, and discussions to [GitHub Issues](https://github.com/CSTR-Edinburgh/merlin/issues).\n\nCitation\n--------\n\nIf you publish work based on Merlin, please cite: \n\nZhizheng Wu, Oliver Watts, Simon King, \"[Merlin: An Open Source Neural Network Speech Synthesis System](https://isca-speech.org/archive/SSW_2016/pdfs/ssw9_PS2-13_Wu.pdf)\" in Proc. 9th ISCA Speech Synthesis Workshop (SSW9), September 2016, Sunnyvale, CA, USA.","github_created_at":"2016-08-10T16:47:10+00:00","created_at":"2026-07-11T12:12:18.671856+00:00","updated_at":"2026-07-12T04:58:01.046368+00:00","categories":[{"slug":"speech-audio","name":"Speech & Audio","url":"https://www.graphcanon.com/categories/speech-audio","markdown_url":"https://www.graphcanon.com/categories/speech-audio.md","api_url":"https://www.graphcanon.com/api/graphcanon/categories/speech-audio"}],"tags":[{"slug":"voice-conversion","name":"voice-conversion"},{"slug":"deep-learning","name":"deep-learning"},{"slug":"text-to-speech","name":"text-to-speech"},{"slug":"keras","name":"keras"},{"slug":"theano","name":"theano"},{"slug":"speech-synthesis","name":"speech-synthesis"},{"slug":"tensorflow","name":"tensorflow"}],"trust":{"provenance":{"is_fork":false,"github_id":65399782,"owner_type":"Organization","methodology":"github_public_v1","parent_repo":null,"near_duplicate_slugs":[]},"computed_at":"2026-07-11T12:12:19.401Z","maintenance":{"label":"Dormant","score":18,"methodology":"github_public_v1","releases_90d":0,"days_since_push":2321,"last_release_at":"2016-11-04T11:59:21Z"},"security_summary":{"status":"findings","scanner":"osv@v1","low_count":594,"high_count":0,"last_scan_at":"2026-07-11T12:12:26.861Z","medium_count":0,"scan_profile":"deps","critical_count":0}},"capability_facts":{"scan":{"source":"repo_scan","observed_at":"2026-07-12T04:58:00.997Z"},"languages":{"value":["python"],"source":"github.language","observed_at":"2026-07-12T04:58:00.997Z"},"license_spdx":{"value":"Apache-2.0","source":"github.license","observed_at":"2026-07-12T04:58:00.997Z"}},"decision_facts":{"hosting":null,"pricing":null,"requirements":{},"constraints":null,"when_to_use":[],"when_not_to_use":[],"source":"enrich:decision_facts","observed_at":"2026-07-12T04:55:46.445Z"},"constraint_facets":null,"decision_summary":[{"label":"Adopt for","value":"Merlin is an open-source Python toolkit for building Deep Neural Network models for statistical parametric speech synthesis that requires pairing with a text processor and vocoder."}]}}