René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"🐸  TTS roadmap","articleBody":"These are the main dev plans for :frog: TTS. \r\n\r\nIf you want to contribute to :frog: TTS and don't know where to start you can pick one here and start with our [Contribution Guideline](https://github.com/coqui-ai/TTS#contribution-guidelines). We're also always here to help. \r\n\r\nFeel free to pick one or suggest a new one. \r\n\r\nContributions are always welcome :muscle:  .\r\n\r\n\r\n## v0.1.0 Milestones\r\n- [x] Better model config handling #21 \r\n- [x] TTS recipes for public datasets.\r\n- [x] TTS trainer API to unify all the model training scripts.\r\n- [x] TTS, Vocoder and SpeakerEncoder model abstractions and APIs. \r\n- [x] Documentation for \r\n   - [x] Implementing a new model using :frog: TTS.\r\n   - [x] Training a model on a new dataset from gecko. \r\n   - [x] Using ```Synthesizer``` interface on ```CLI``` or ```Server```.\r\n   - [x] Extracting Spectrograms for Vocoder training. \r\n   - [x] Contributing a new pre-trained :frog: TTS model.\r\n   - [x] Explanation for Model config parameters/\r\n\r\n## v0.2.0 Milestones\r\n- [x] Grapheme 2 Phoneme in-house conversion. (Thx to gruut  👍 )\r\n- [x] Implement VITS model.\r\n\r\n## v0.3.0 Milestones\r\n- [x] Implement generic ForwardTTS API.\r\n- [x] Implement Fast Speech model. \r\n- [x] Implement Fast Pitch model. \r\n\r\n## v0.4.0 Milestones\r\n- [x] Trainer API v2 - [join the discussion](https://github.com/coqui-ai/TTS/discussions/840)\r\n- [x] Multi-speaker VCTK recipes for all the `TTS.tts` models.\r\n\r\n## v0.5.0 Milestones\r\n- [x] Support for multi-lingual models \r\n- [x] YourTTS release 🚀\r\n\r\n## v0.6.0 Milestones\r\n- [x] Add ESpeak support\r\n- [x] New Tokenizer and Phonemizer APIs #937 \r\n- [x] New Model API #1078 \r\n- [x] Splitting the trainer as a separate repo [👟Trainer](https://github.com/coqui-ai/Trainer)\r\n- [x] Update VITS model API\r\n- [x] Gradient accumulation. #560 (in 👟)\r\n\r\n## v0.7.0 Milestones\r\n- [x] Implement Capacitron 👑 @a-froghyar 👑 @WeberJulian \r\n- [x] Release pretrained Capacitron\r\n\r\n## v0.8.0 Milestones\r\n- [x] Separate numpy transforms \r\n- [x] Better data sampling for VITS \r\n- [x] New Thorsten DE models 👑 @thorstenMueller\r\n\r\n## 🏃‍♀️ Milestones along the way\r\n- [ ] Implement End-to-end training API for ForwardTTS models a vocoder. #1510 \r\n- [ ] Implement a Python voice synthesis API.\r\n- [ ] Inject phonemes to the input text at inference. #1452 \r\n- [ ] AdaSpeech1/2 https://arxiv.org/pdf/2104.09715 and https://arxiv.org/abs/2103.00993\r\n- [ ] Let the user pass a custom text cleaner function.  \r\n- [ ] Refactor the text cleaners for a more flexible and transparent API. \r\n- [ ] Implement [HifiGAN2](https://pixl.cs.princeton.edu/pubs/Su_2021_HSS/Su-HiFi-GAN-2-WASPAA-2021.pdf) (not the vocoder) \r\n- [ ] Implement emotion and style adaptation. \r\n- [ ] Implement FastSpeech2 (https://arxiv.org/abs/2006.04558). \r\n- [ ] [AutoTTS 🤖](https://github.com/coqui-ai/TTS/pull/696) (👑 @loganhart420)\r\n- [ ] Watermarking TTS outputs to sign against DeepFakes.\r\n- [ ] Implement SSML v0.0.1 \r\n- [ ] ONNX and TorchScript model exports. \r\n- [ ] TensorFlow run-time for training models.\r\n\r\n## 🤖 New TTS models \r\n- [x] AlignTTS (@erogol) \r\n- [x] HiFiGAN (#16 :crown: @rishikksh20 and @erogol)\r\n- [x] UnivNet Vocoder ( :crown: @rishikksh20)\r\n- [x] VITS [paper](https://arxiv.org/abs/2106.06103)\r\n- [x] FastPitch [source](https://fastpitch.github.io/)\r\n- [x] Alignment Network [paper](https://arxiv.org/abs/2108.10447) \r\n- [x] End2End TTS combining aligner + tts + vocoder. \r\n- [x] Multi-Lingual TTS (#11 :crown: @WeberJulian )\r\n- [ ] ParallelTacotron [paper](https://arxiv.org/pdf/2103.14574.pdf) (open for contribution)\r\n- [ ] Efficient TTS [paper](https://arxiv.org/pdf/2012.03500.pdf) (open for contribution)\r\n- [ ] Gaussian length regulator from https://arxiv.org/pdf/2010.04301.pdf (open for contribution)\r\n- [ ] LightSpeech from https://arxiv.org/pdf/2102.04040.pdf (open for contribution)\r\n- [ ] AdaSpeech1/2 https://arxiv.org/pdf/2104.09715 and https://arxiv.org/abs/2103.00993","author":{"url":"https://github.com/erogol","@type":"Person","name":"erogol"},"datePublished":"2021-03-13T14:14:51.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":62},"url":"https://github.com/378/TTS/issues/378"}