Skip to main content
Speech recognition

Speech recognition

Type of softwareWikipedia

Search complete. 30 mentions across 8 episodes found for "Speech recognition".

Nov 22, 2026

speaker_0HOST
0:59
He's got speech in his demos.
speaker_0HOST
1:01
He's got a Liquendo TTS and ASR engine, and he just told me that his ASR license is having problems.
speaker_0HOST
1:07
So you'll have to take his ASR bits on faith, but hopefully the TTS bits will work in this demo.
speaker_0HOST
1:14
So take it from there.
speaker_0HOST
1:15
And we are going to be recording this talk, so hold questions for after the video cameras have stopped rolling, since we'll be putting this up on Google Video.

22 MINS LATER

Dave RaggettGUEST
23:41
So it's all standard stuff.
Dave RaggettGUEST
23:42
Using HTTP with the payloads as a MIME, you know, internet media types for the payloads for the audio, PCM audio.
Dave RaggettGUEST
23:49
With the ASR and the HTTP request for speech recognition.
Artificial IntelligenceNARRATOR
4:50
Immersive VR Conversations with iAvatars.
Artificial IntelligenceNARRATOR
4:54
Integrating Chad GPT, Google STT, and AWS Polly learn how integrating Chad GPT, Google Speech-to-Text, and Amazon Web Services Polly in VR can create realistic and interactive conversations with AI avatars.
Artificial IntelligenceNARRATOR
5:09
18.
Artificial IntelligenceNARRATOR
5:09
Why the beauty and fashion industry is the future of AR and SPATIALCOMPUTINGAR and VR finally have their killer app, the beauty and fashion industry.
Hongyi LiHOST
89:50
This question is, you identify noisy input channels such as speech as a source of environmental uncertainty.
Hongyi LiHOST
90:01
How should ASR uncertainty be passed into the agent's decision policy rather than treating the Transcripts as ground truth.
Hongyi LiHOST
90:15
Is this harder for a typical speech such as children's or essence speech where recognition errors are systematic rather than random?
Kai YuGUEST
90:29
Okay.
Kai YuGUEST
90:30
I think this is a very good question.
Kai YuGUEST
90:32
I can tell you some of our previous research on this.
Kai YuGUEST
90:38
So if you pass just one best hypothesis of ASR system, that's one option.
Kai YuGUEST
90:45
That's a normal option to a large language model.
Alston GodboltHOST
1:28
I've got my Google Docs open.
Alston GodboltHOST
1:29
I'm going to go to Tools, and I'm going to go to Voice Typing.
Alston GodboltHOST
1:35
And this is what I'm going to do.
Alston GodboltHOST
1:36
I'm just going to talk to it, okay? And the reason why I'm using Voice Typing is so I can copy and paste what I say into Cloud.
Alston GodboltHOST
1:44
So watch this.
Alston GodboltHOST
1:46
I want to start an online business, but I'm not really sure where to go.
Harrison KinsleyHOST
25:30
To make it full, fully, um, you know, end to end.
Harrison KinsleyHOST
25:33
So here we can see they're kind of talking here, but I think ASR that's automatic or automated speech recognition, essentially speech to text.
Harrison KinsleyHOST
25:43
Uh, there's other, it's not just speech to text text, which is why I think they're using ASR here.
Harrison KinsleyHOST
25:49
There's a whole lot more going on there.
Harrison KinsleyHOST
25:51
And, um, there's a live microphone demo that I'm going to try to do by the end of this video.
Harrison KinsleyHOST
26:44
Okay, so in this case, we can see here, okay, they're doing some NLP here.
Harrison KinsleyHOST
26:48
This is probably just understanding of sentences generally with NLP.
Harrison KinsleyHOST
26:53
ASR, this is probably mostly speech to text, but we can kind of dig into it in a little bit.
Linda WilsonHOST
0:43
There was a time when a transcriptionist would sit down with an audio recording and type practically every word from scratch.
Linda WilsonHOST
0:52
That work still exists, but we're also seeing more companies use automatic speech recognition, also called ASR, to create the first rough draft of a transcript, and that changes the job.
Linda WilsonHOST
1:08
So instead of starting with a blank page and typing every word, a trained editor may start with a computer-generated rough draft and work from there.
Linda WilsonHOST
1:19
But now here's the important part.
Linda WilsonHOST
2:27
The work may simply move in a different direction.
Linda WilsonHOST
2:30
We already see digital reporters capturing legal proceedings while other professionals prepare the written transcript.
Linda WilsonHOST
2:40
Increasingly, that transcript may begin with an ASR-generated rough draft rather than a blank page.
Linda WilsonHOST
2:48
That creates another kind of skill set legal transcript editing.
Andrew KatzGUEST
17:04
or models in this case.
Andrew KatzGUEST
17:06
They'll also go by the phrase of automatic speech recognition.
Andrew KatzGUEST
17:09
So you'll either see the acronym TTS, sorry, STT, so speech to text.
Andrew KatzGUEST
17:15
You'll also see TTS, that's text to speech.
Andrew KatzGUEST
17:18
So there's the STT kinds of tools, but that also goes by the acronym ASR, so that's just the automatic speech recognition.
MoniqueHOST
17:25
The most
Andrew KatzGUEST
17:26
popular one for a number of years was actually, coincidentally, from OpenAI, but this was back when they were actually pushing out open source models.
Samantha JoyHOST
10:52
You can use a computer to type this out, but you can also go into Google Docs, for example.
Samantha JoyHOST
10:56
I think Word does this as well, where you can go into Tools in the menu, and it's called Voice Typing.
Samantha JoyHOST
11:03
And you can close your eyes and just speak to your computer.
Samantha JoyHOST
11:06
You may need to say period, exclamation point, quotation.

We value your privacy

We use cookies to understand how you use our platform and to improve your experience. Click “Accept All” to consent, or “Decline non-essential” to opt out of non-essential cookies. Read our Privacy Policy.