Okay, I visited my university, Stockholm University. Just to see the companies working on speech recognition software in some form. I spoke to scientist and promotion staff. Most of them where white asian men, even though there where some asian women, some white men and a few white women. I spoke and interacted with at least 50% of the companies there. I asked them about their products and listened. I then asked why they made the product or that research. I don’t think they where expecting that question. I wanted to know the purpose for them doing the research. It seems as a fair question. I am guessing some of them don’t even know. So I had to say what it could be used for. Sometimes I even said what the negative things of their research could result in. They didn’t disigree in any case, since they know it’s possible. It is as if the research or product has a Devil as well as God inside by default. Bad and good at the same time.
One product was really good and the researcher had come so far in his research in just a year. I was impressed. I liked talking to him. He was good at explaining a complex technology. He had thought it threw and he knew how to keep it explainable, yet still complex. He had developed a system that fills in the gaps for people missing a tounge. Instead of using a real proteses the computer interprets and gives a real speech from the persons speak. It was a overall good improvement. Some disturbance in noise, but then again he had made great progress in just one year. I didn’t tell him it could be used to fill in the gaps in surveillance phone calls… but it can. When we cannot hear all… I think that is a possible development of that software. His research is good for people who have lost their toungue. I have a soft spot for health products in general. Goodness. But evilness can always twist even those products.
Then there are systems created to recognise voices in crowds, to spot a specific voice in a crowd. Imagine what that could be used for. Do I need to explain? To identify you of course. Imagine war driving with these kinds of features. Haha. Instead of tapping into wireless networks, they could collect voices and find conversations from a far distance with a automated system, so you only get a transcript with another system by that audio file. Then there where systems that focused on how to pack the audio files and making them comparable. Comparison gives totally new possibilities. They explained how they did it and it was interesting to hear different solutions.
One product was a about designing voices. That could be used for future robots, so they get humanlike voices. Then your future robot housekeeper can sound like your grandmother. You only need 10 sentences with your granny to catch her voice. Imagine that for scams. Catching someones voice in public, then doing scams making phone calls as that person by a system represented there. Well. Then totally automated customer services, for customer interaction at big sales markets online. Then add cars that identify the user and adapts all features to you in the car, then your voice can command air, measurement or any other feature you wanna use and call up on in the car, disregarding others, if that is your wish. Well. Then there where those focused on social media. Snap chat even came up. So well… they believed the future was audio… not images. My teacher’s have been saying they will be visual. But some think that speech will be even more at this mass. Hmm… well, Are you totally satisfied with voice recognition? Are you? How many times haven’t we been subject to misinterpretation? Many times… haha. Google maps… fucking improve them.
Well. You should have gotten this information. Someone should present all these things online. It is pretty interesting to hear all the future products out there. The products are in many cases really real in research already.
The event was called Interspeech. It is going on tomorrow too.