Technology has both sides to it, one is good while the other is bad. And we know that we have to be updated with technology because it is going to take over our lives in the future. However, there are many who feel that too much technology is not good for humans. Still, we have to keep pace with technology and learn to not get too much of technology in our lives.
You must have heard about how technology can accurately predict when you will get a heart attack or even when you will die.
Also, there are some procedures with which you can predict how someone sounds based on their look. However, we now have an algorithm which predicts how a person looks based on their sound. The idea is excellent and you would think that the results would be interesting.
Well, you would be surprised to know that the algorithm’s portrait was very very accurate to the level of being fearful. This algorithm has been developed by students at the MIT who are doing research on a paper called Speech2Face: Learning the Face Behind a Voice. This algorithm will draw a portrait of the speaker based on a short recording of just the speaker’s voice.
Now, it must be said that the portraits were not exactly the same but they were able to accurately predict the gender, race, and age as well as the face of a speaker. It is known that the deep neural network was made to understand more than 100,000 different speakers based on their YouTube clips.
The researchers do note, however, that the neural network does not reveal “the true identity of a person” but only shows “average-looking faces” which give you an idea of how the person looks. The researchers say that this is done to not intrude someone’s privacy.