31–40 of 172 posts
Slightly OT, but in school I remember a teach talking about how good voice to text has gotten but was stuck, stating it was roughly about a 70% accuracy, but that we were pretty much at a wall there and haven't moved much in the last decade. Has there been any major advances in voice recognition other than just growing your speech corpus?
"What's going to be on my ballot?"
Can Watson even answer this?
For a good believable test your sampling methodology is very important. In fact, you want to have a random sample that is free of biases instead of somebody cherry picking queries. Perhaps you may do random weighted sample of queries with weights = frequency which represents usage pattern more closely. In any case, it's very important to describe your sampling methodology or otherwise this kind of testing has little value.
I think this analysis is kind of pointless. Who asks their phone trivia questions? You want directions, reminders, etc. Siri seems a lot smarter about following spoken context in that situation.
Oh awesome, let's see... how many atoms are there in the universe? "The number of atoms in the entire observable universe is estimated to be within the range of 1078 to 1082." Thank you Google! http://www.google.com/search?q=how+many+atoms+are+in+the+uni... ]
It's also worth noting that Siri gives more verbose answers in "Hey Siri" mode, presumably because it assumes you're not looking at the screen.
Who owns the key patents for voice recognition - AT&T, Nuance, IBM? When will they be expire and be available for use in open-source speech recognition? Edit: some data at http://www.quora.com/Is-anyone-working-on-an-open-source-ver... & https://news.ycombinator.com/item?id=4987875
Move along, nothing to see here.
Oh awesome, let's see... how many atoms are there in the universe? "The number of atoms in the entire observable universe is estimated to be within the range of 1078 to 1082." Thank you Google! http://www.google.com/search?q=how+many+atoms+are+in+the+uni... ]
10^78 and 10^82 Just a formatting issue.
Solving it isn't really easy, either. When reading a document is a caret a power, a regular expression operator, an exclusive or, or a nose on a smiley? :^)
A lot of context identification work has to be sorted out for smart agents to succeed.
I think this analysis is kind of pointless. Who asks their phone trivia questions? You want directions, reminders, etc. Siri seems a lot smarter about following spoken context in that situation.
I can't speak extensively for Siri (infrequent use on my iPad) or Google Now (no Android), but I've never had to repeat an instruction to Cortana for a reminder or directions. Not repeating myself when giving voice instruction is essentially my holy grail.
Oh awesome, let's see... how many atoms are there in the universe? "The number of atoms in the entire observable universe is estimated to be within the range of 1078 to 1082." Thank you Google! http://www.google.com/search?q=how+many+atoms+are+in+the+uni... ]
Try it on Wolfram Alpha. Actually, they all seem to me pretty limited compared to the answers you get from Wolfram Alpha. Also c.f. the responses to one of their test questions, "how much is a quarter cup of butter?". Google makes fun of the inquiry. Wolfram Alpha gives you a thorough nutritional profile, and links to variations based on international cup sizing and different types of butter.
In the article, the question "How old is the Lincoln Tunnel" struck me as incorrectly formatted for the parser (I know, that's the point), so I asked Siri, "When was the Lincoln Tunnel built." The Wikipedia article on the Lincoln Tunnel was returned. Wolfram Alpha was listed under other sources, so I chose that. The response? "1937"
