Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Updates to Cloud Speech-to-Text and general availability of Cloud Text-to-Speech

cloud.google.com

Post reply on HN
Print view View original

21–30 of 42 posts

Re: Updates to Cloud Speech-to-Text and general availability of Cloud Text-to-Speech

#21

Post by dvfjsdhgfv »

pasta wrote: post #19

A friend is working for a newspaper. He records interviews. We tried all the software we could find to turn the recording (Dutch) into text but there is nothing that gives a helpful result. I know that a recording-to-text is different than speech-to-text but even when I use OK Google most of the time the results are horrible. So after all those years I am still a little skeptical.

That's why all the hype about DL is so irritating. Yes, it allowed us to make enormous progress in certain fields. No, it's nowhere near usable in others. So why pretend it is?

Just one example from the preface to Chollet's "Deep Learning with Python":

> If you’ve picked up this book, you’re probably aware of the extraordinary progress that deep learning has represented for the field of artificial intelligence in the recent past. In a mere five years, we’ve gone from near-unusable image recognition and speech transcription, to superhuman performance on these tasks.

Come on, speech to text is still far from usable unless in a very limited scenarios. Why pretend it's different?

Re: Updates to Cloud Speech-to-Text and general availability of Cloud Text-to-Speech

#22

Post by dan0- »

singularity2001 wrote: post #18
moconnor wrote: post #11

Earlier quoted context omitted.

Mozilla’s DeepSpeech is available as an offline pre-trained system for English. The quality is far below Google’s speech API as the model is somewhat out of date and more importantly the training data set is much smaller and less general. The best pretrained speech to text model I’ve seen is from Baidu’s DeepSpeech 2 repository. They provide pretrained models for English and Chinese based on their internal data. The…

> can be comfortably run in real-time on a desktop that's progress. how long did it take you to get these things running (including downloads and dependency installation)?

The downloads total a bit above 2GB for Mozilla Deepspeech 0.1.1, but besides that DeepSpeech is quick to set up and pretty performant on my i5-4200U (half realtime transcription) and its even better on my Ryzen box. Going to do some GPU testing tomorrow, need to clean up the docs on the little web frontend/API I've been trying to get ready for production use: https://github.com/AccelerateNetworks/DeepSpeech_Frontend

Re: Updates to Cloud Speech-to-Text and general availability of Cloud Text-to-Speech

#23

Post by dan0- »

danShumway wrote: post #12
moconnor wrote: post #11

Earlier quoted context omitted.

Mozilla’s DeepSpeech is available as an offline pre-trained system for English. The quality is far below Google’s speech API as the model is somewhat out of date and more importantly the training data set is much smaller and less general. The best pretrained speech to text model I’ve seen is from Baidu’s DeepSpeech 2 repository. They provide pretrained models for English and Chinese based on their internal data. The…

I'm happy to be wrong about that; I was under the impression that DeepSpeech was available as a data set for training , but not as a trained system in and of itself. My impression is that a lot of the hardware costs from modern AI comes from training and updating the model. So (speaking as a non-expert) it doesn't seem like there's any technological problem with running a model locally on a private network -- you jus…

DeepSpeech is not a voice corpus, that would be Mozilla's Common Voice project that Mycroft users are also contributing to.

DeepSpeech offers trained models that are about 70% right, but none of them use the Common Voice corpus yet. I think that plus recent changes to the codebase should produce a much better transcription, but I don't have the GPU resources to go and train a model sadly, Mozilla will hopefully release another trained model soon though!

I am working on a basic web frontend and API for DeepSpeech: https://github.com/AccelerateNetworks/DeepSpeech_Frontend

Also, here is Common Voice: https://voice.mozilla.org/en

Re: Updates to Cloud Speech-to-Text and general availability of Cloud Text-to-Speech

#24

Post by anonymous5133 »

saas_co_de wrote: post #20
danShumway wrote: post #10

Google's speech-to-text is powerful, but I'd be pretty skeptical about tying a project to it given how services like Maps have been handled recently. There are companies like Mozilla trying to build more open solutions, but to the best of my knowledge (please correct me if I'm wrong) any pre-trained services Mozilla offers will also still involve you connecting to their servers. Maybe I'm just paranoid, but I just ca…

It is also extremely expensive ($1.44/hr of audio) which really limits its potential. I looked at it for a project but it just didn't make sense.

I've been using it for one of my projects but it just requires one-time use of the audio. As in, I use the text to speech to create audio files which I just recycle. Overall, it is more effective compared to using my own voice or hiring a voice over person to do it.

Re: Updates to Cloud Speech-to-Text and general availability of Cloud Text-to-Speech

#25

Post by singularity2001 »

dan0- wrote: post #22
singularity2001 wrote: post #18

Earlier quoted context omitted.

> can be comfortably run in real-time on a desktop that's progress. how long did it take you to get these things running (including downloads and dependency installation)?

The downloads total a bit above 2GB for Mozilla Deepspeech 0.1.1, but besides that DeepSpeech is quick to set up and pretty performant on my i5-4200U (half realtime transcription) and its even better on my Ryzen box. Going to do some GPU testing tomorrow, need to clean up the docs on the little web frontend/API I've been trying to get ready for production use: https://github.com/AccelerateNetworks/DeepSpeech_Frontend

I was more interested in 'astonishingly good' Baidu solution. will that install chinese spyware thou? sorry, very cautious about Baidu et al.

Re: Updates to Cloud Speech-to-Text and general availability of Cloud Text-to-Speech

#26

Post by dan0- »

moconnor wrote: post #11
danShumway wrote: post #10

Google's speech-to-text is powerful, but I'd be pretty skeptical about tying a project to it given how services like Maps have been handled recently. There are companies like Mozilla trying to build more open solutions, but to the best of my knowledge (please correct me if I'm wrong) any pre-trained services Mozilla offers will also still involve you connecting to their servers. Maybe I'm just paranoid, but I just ca…

Mozilla’s DeepSpeech is available as an offline pre-trained system for English. The quality is far below Google’s speech API as the model is somewhat out of date and more importantly the training data set is much smaller and less general. The best pretrained speech to text model I’ve seen is from Baidu’s DeepSpeech 2 repository. They provide pretrained models for English and Chinese based on their internal data. The…

Any links to Baidu's DeepSpeech 2 pretrained model? I'm likely looking right past it, spent the past few days playing around with Mozilla DeepSpeech and their pretrained model, building a simple little API and webpags to feed it files: https://github.com/AccelerateNetworks/DeepSpeech_Frontend

I really want to put DeepSpeech 2 with Baidu's model up against Mozilla's model and see which is better, seems like it could be quite interesting!

Re: Updates to Cloud Speech-to-Text and general availability of Cloud Text-to-Speech

#27

Post by dan0- »

singularity2001 wrote: post #25
dan0- wrote: post #22

Earlier quoted context omitted.

The downloads total a bit above 2GB for Mozilla Deepspeech 0.1.1, but besides that DeepSpeech is quick to set up and pretty performant on my i5-4200U (half realtime transcription) and its even better on my Ryzen box. Going to do some GPU testing tomorrow, need to clean up the docs on the little web frontend/API I've been trying to get ready for production use: https://github.com/AccelerateNetworks/DeepSpeech_Frontend

I was more interested in 'astonishingly good' Baidu solution. will that install chinese spyware thou? sorry, very cautious about Baidu et al.

Ah, I just asked a similar question of the person you initially replied to. The Baidu DeepSpeech 2 model seems appealing, but from my brief look on Github I haven't found the model files...

Re: Updates to Cloud Speech-to-Text and general availability of Cloud Text-to-Speech

#28

Post by singularity2001 »

dan0- wrote: post #23
danShumway wrote: post #12

Earlier quoted context omitted.

I'm happy to be wrong about that; I was under the impression that DeepSpeech was available as a data set for training , but not as a trained system in and of itself. My impression is that a lot of the hardware costs from modern AI comes from training and updating the model. So (speaking as a non-expert) it doesn't seem like there's any technological problem with running a model locally on a private network -- you jus…

DeepSpeech is not a voice corpus, that would be Mozilla's Common Voice project that Mycroft users are also contributing to. DeepSpeech offers trained models that are about 70% right, but none of them use the Common Voice corpus yet. I think that plus recent changes to the codebase should produce a much better transcription, but I don't have the GPU resources to go and train a model sadly, Mozilla will hopefully relea…

Keep up the good work ++

Re: Updates to Cloud Speech-to-Text and general availability of Cloud Text-to-Speech

#29

Post by singularity2001 »

dvfjsdhgfv wrote: post #21
pasta wrote: post #19

A friend is working for a newspaper. He records interviews. We tried all the software we could find to turn the recording (Dutch) into text but there is nothing that gives a helpful result. I know that a recording-to-text is different than speech-to-text but even when I use OK Google most of the time the results are horrible. So after all those years I am still a little skeptical.

That's why all the hype about DL is so irritating. Yes, it allowed us to make enormous progress in certain fields. No, it's nowhere near usable in others. So why pretend it is? Just one example from the preface to Chollet's "Deep Learning with Python": > If you’ve picked up this book, you’re probably aware of the extraordinary progress that deep learning has represented for the field of artificial intelligence in the…

in most languages google ASR works shockingly well.

Re: Updates to Cloud Speech-to-Text and general availability of Cloud Text-to-Speech

#30

Post by singularity2001 »

zawerf wrote: post #15

Anyone know how this relates to the Web Speech API[1]? Will they ship it with chrome to replace the existing speech synthesis api? (I believe right now it just uses whatever voices are available to the device or OS but chrome can fallback to a serverside voice) [1] https://developer.mozilla.org/en-US/docs/Web/API/Web_Speech_... [2] https://developer.mozilla.org/en-US/docs/Web/API/SpeechSynth...

pure guess: Mozilla will use their DeepSpeech model for that soon
Post reply on HN