3 ms·

Does an offline voice assistant fit on a watch or a smart phone? Is it possible to open source one?

• albertzeyer 7y ago
  Yes. The big parts are the models (neural networks (NN)). You can make them quite small (sparsification, quantization, transfer learning to smaller models, ...). But this is non-trivial. Kaldi is the most complete open source speech recognition software, covering HMM/GMMs, the conventional hybrid HMM/NN, and much more. If you care about end-to-end only, there is much more, e.g. Mozilla DeepSpeech, and most modern open source encoder-decoder-attention frameworks can do that (e.g. Lingvo etc). Most of these encoder-decoder-attention framework can also do text-to-speech (TTS). Maybe you need a separate vocoder, but you will also find open source code for that. For the assistant engine, there are also a couple of open source projects, e.g. Mycroft, LinTO, Rasa, Leon, Open Assistant, Susi, ... Putting it all together is non-trivial, and requires lots of expert knowledge, and lots of work and time to get a good result. This is hard to do as a hobby project.
• dest 7y ago
  Have a look at https://snips.ai/ https://snips.ai/
  • VistaBrokeMyPC 7y ago
    That site has the most GDPR compliant, transparent, easy to configure cookie settings I've ever seen. I've got my eyes on that dev kit for sure. Having an offline voice assistant is something I've wanted for a while, privacy is much more important to me than the hardware support that echo and Google home have. Thanks for dropping that link!