-
- AI News
- [페이스북, 뉴욕대 연구 계정 차단…데이터 연구활용 놓고 갈등](https://n.news.naver.com/mnews/article/001/0012579462?sid=105)
- [DEVIEW 2021 연사 모집](https://deview.kr/2021/cfs)
- 8월 25일까지
- 이번에도 온라인으로…
-
I'm trying to train a FastSpeech2 model on a dataset i scraped myself (around 7 hours of speech and 5000 audio files resampled to 22050hz). I do not expect great results, but at least something to wor…
-
The positional encoding used in the code is
https://github.com/janvainer/speedyspeech/blob/3a30645249c9f60b9655626afbb263f3e43a4758/code/functional.py#L43
I think it is intended to use the posit…
-
In #126 it is mentioned that most of the ability to clone voices lies in the encoder. @mbdash is contributing a GPU to help train a better encoder model.
* Increase the number of hidden layers to 7…
ghost updated
3 years ago
-
Below are Colabs notebooks for Demo of two models, the demos allow upload of GST samples, upload of wav file to use as a reference for the speaker (synthesize with your own voice).
[Demo Mozilla TT…
-
Bonjour,
J'aimerai savoir comment faire pour ajouter/modifier une voix de svox par exemple celle de Luc ?
Les voix de robot ne sont pas très atrayantes :)
Merci
-
I have been training MelGAN-STFT by finetuning it on the LJSpeech model. When it gets to `discriminator_train_start_steps`, it stops and tells me to restart. When I restart with the discriminator on l…
-
When publishing to `hermes/dialogueManager/endSession` it looks like the text field defined in the [spec](https://docs.snips.ai/reference/dialogue#outbound-message-2) isn't respected (i.e., doesn't le…
-
Global Style Tokens are embeddings that capture prosodic styles across the training set. This allows the system to explicitly specify the desired prosody of a generated sequence, i.e. essentially how …
-
Hi @erogol, I've been a bit off the radar for the past month because of vacation and other projects, but now I am back and ready for action! I am looking into how to do multi speaker embeddings, and h…