25 Years In Speech Technology and I still don’t talk to my computer
matthewkaras.medium.com
25 Years In Speech Technology and I still don’t talk to my computer
1–10 of 296 posts
Re: 25 Years In Speech Technology and I still don’t talk to my computer
#2Very unlikely I will ever talk to my computer irrespective of how good the speed technology gets. If I have to talk to my computer how will I work in crowded places ? How does it work ?
Re: 25 Years In Speech Technology and I still don’t talk to my computer
#3Very unlikely I will ever talk to my computer irrespective of how good the speed technology gets. If I have to talk to my computer how will I work in crowded places ? How does it work ?
Wear some sort of sound containing bubble around your head, fits in nicely with the whole pandemic thing :)
Re: 25 Years In Speech Technology and I still don’t talk to my computer
#4Very unlikely I will ever talk to my computer irrespective of how good the speed technology gets. If I have to talk to my computer how will I work in crowded places ? How does it work ?
In a croweded place the noise has to be cancelled or speech recognition application must learn to recognise with noise in input audio.
Re: 25 Years In Speech Technology and I still don’t talk to my computer
#5A big problem with the assistants is that as soon as they fail at a query they seem stupid, I feel stupid, and I stop using them for a long time.
Last weekend I had the following failed queries:
"OK Google, what is the air quality like at Mt. Shasta today?"
"OK Google, add a waypoint for the last gas station before the mountain pass"
"OK Google, what percentage of people can you detect to be wearing masks on recent Instagram photos tagged at a location within a 10 mile radius of Mt. Shasta?"
These are all things I would expect a computer assistant to do really well. They have access to so much data, and so many APIs, that they should be able to break down these sentences into a SQL-like query and give me results. The third, for example:
"recent Instagram photos" -> Instagram has an API
"tagged within a 10 mile radius" -> parse the cities within a 10 mile radius and look for tags in all of them
"people" -> use your wonderful person detection networks your friends at Waymo developed
"wearing masks" -> I'm sure your internal datasets have this label, so run an object detector
Then compile and reduce the data to give me the number I want.
That's what I want an assistant to do. But it couldn't even do the first, which just involves a single API query to fetch air quality index information. Bleh. And as for the second, it has no idea what "last gas station" or "mountain pass" means; it's a query a human would know to be extremely commonplace.
It turns out that the current generation of "assistants" are mostly just template-matchers which really doesn't help me much at all. I can set my own alarms, thank you.
Re: 25 Years In Speech Technology and I still don’t talk to my computer
#6Very unlikely I will ever talk to my computer irrespective of how good the speed technology gets. If I have to talk to my computer how will I work in crowded places ? How does it work ?
In a croweded place the noise has to be cancelled or speech recognition application must learn to recognise with noise in input audio.
I think the real problem is social, no one wants to hear you talking at your computer in a coffee shop.
Re: 25 Years In Speech Technology and I still don’t talk to my computer
#7I will consider talking to my computer when my computer is the only one that would hear me.
I also never talk to my phone.
Re: 25 Years In Speech Technology and I still don’t talk to my computer
#8Very unlikely I will ever talk to my computer irrespective of how good the speed technology gets. If I have to talk to my computer how will I work in crowded places ? How does it work ?
In a croweded place the noise has to be cancelled or speech recognition application must learn to recognise with noise in input audio.
I think part of the issue is that we're pretty good at focusing and filtering. If we're trying, we can focus on a single voice when the room is noisy. The computer doesn't know which to choose and all voices seem equally important.
Re: 25 Years In Speech Technology and I still don’t talk to my computer
#9I dearly wish I could have a Speech Recognition aspect to the 'creative' and CAD-like software I use - I think it would be fantastic to be able to do two or three operations 'at the same time' - eg. click an object with the mouse, hit ctrl on the keyboard to lock to some plain or angle, as I use the mouse to move, or whatever, whilst verbally instructing the software to 'zoom viewport out' or achieve whatever other function that is vaguely complicated or buried in the GUI somewhere.
Otherwise, I have absolutely no desire to 'talk' to my computer and have it understand me, unless that tech comes packaged with an empathic AI module so I can tell it off and repay it a small percentage of the emotional pain computers have inflicted upon me over the decades.
Re: 25 Years In Speech Technology and I still don’t talk to my computer
#10There are good use cases apart from Dictation for Speech tech:
1. Voice based operations in factories, construction workers who want to have both the hands free but want to navigate via a device
2. Use cases while Driving. E.g. A Driver who is delivering goods.
3. Call centre - Analytics of audio calls etc. Can have many use cases
4. Voice assistant like Alexa, Siri. Mind that Alexa, Siri have vision to do more than just Music.
5. Any use case where visual interface is either not there or visual is not an easy option for user.
Speech tech is challenging when you have to deal with noise or want to do Speech to text on lw profile devices (on the edge).