This all doesn’t mean that women’s voices are more difficult I’ve trained classifiers on speech data from women and they worked just fine, thank you very much. Manning where they found better performance for women than men–due to the above factors and different rates of filler words like “um” and “uh”.) One thing that may be making a difference is that women also tend not to be as loud, partly as a function of just being smaller, and cepstrals (the fancy math thing what’s under the hood of most automatic voice recognition) are sensitive to differences in intensity. (Edit : I have since found two papers by Sharon Goldwater, Dan Jurafsky and Christopher D. Women’s vowels, in particular, lend themselves to classification: women produce longer vowels which are more distinct from each other than men’s are. In general, women also favor more standard forms and make less use of stigmatized variants. Women tend to be more intelligible (for people without high-frequency hearing loss), and to talk slightly more slowly. The suggestion that, for example, “ women could be taught to speak louder, and direct their voices towards the microphone” is ridiculous. In fact, women use speech strategies that should make it easier for voice recognition technology to work on women’s voices. So where is this imbalance coming from? First, let me make one thing clear: the problem is not with how women talk. And that’s not even touching on the safety implications of voice recognition in cars. Even if it only takes a second to correct an error, those seconds add up over the days and weeks to a major time sink, time your male colleagues aren’t wasting messing with technology. The fact that men enjoy better performance than women with these technologies means that it’s harder for women to do their jobs.
GOOGLE SPEECH SOFTWARE
But some of these applications have a lot higher stakes. Take the medical dictation software study. Sure, a few incorrect Youtube captions aren’t a matter of life and death. This is a real problem with real impacts on people’s lives. Paper finding that speech recognition performs worse for women than men, and worse for girls than boys (Nicol et al.Study finding that medical voice-dictation software performs significantly better for men (Roger & Pendharkar 2003).It’s Not You, It’s It: Voice Recognition Doesn’t Recognize Women (Times, 2011).There’s a long history of speech recognition technology performing better for men than women: What it is not, unfortunately, is shocking. The Cohen’s d was 0.7 which means, in non-math-speak, that if you pick a random man and random woman from my sample, there’s an almost 70% chance the transcriptions will be more accurate for the man. It’s not that there’s a consistent but small effect size, either, 13% is a pretty big effect. The average male speaker, on the other hand, was captioned correctly 60% of the time. (You can see my data and analysis here.) On average, for each female speaker less than half (47%) her words were captioned correctly. And when I compared performance on male and female talkers, I found something deeply disturbing: YouTube’s auto captions consistently performed better on male voices than female voice (t(47) = -2.7, p < 0.01.). Now, because I’m a sociolinguist and I know that it’s important to stratify your samples, I made sure I had an equal number of male and female speakers for each dialect. To get this data, I hand-checked annotations more than 1500 words from fifty different accent tag videos. In my last post, I looked at how Google’s automatic speech recognition worked with different dialects. The original, unedited blog post, continues below.
GOOGLE SPEECH PROFESSIONAL
In my professional opinion, the racial differences are both more important and difficult to solve. More recent research has found the same effect: ASR systems make more errors for Black speakers than white speakers. In my 2017 study, I found that multiple commercial ASR systems had higher error rates for non-white speakers. That said, bias against specific demographics categories in automatic speech recognition is a problem. I take this to be evidence that differences in gender are due to differences in overall signal-to-noise ratio when recording in noisy environments rather than problems in the underlying ML models. I’m no longer working actively on this topic, but in the last paper I wrote on it, in 2017, I found that when audio quality was controlled the gender effects disappeared.
GOOGLE SPEECH UPDATE
Edit, July 2020: Hello! This blog post has been cited quite a bit recently so I thought I’d update it with the more recent reserach.
0 kommentar(er)
