![]() The researchers also showed the participants short Pixar videos that didn’t contain any dialogue, and recorded their brain responses in a separate experiment designed to test whether the decoder was able to recover the general content of what the user was watching. For example, a user heard the words “I don’t have my driver’s license yet.” The decoder returned the sentence “She has not even started to learn to drive yet.” When they tested the model on new podcast episodes, it was able to recover the gist of what users were hearing just from their brain activity, often identifying exact words and phrases. It predicted how the brain would respond to the guessed words, and then compared that with the actual measured brain responses. To decode, it guessed sequences of words and checked how closely that guess resembled the actual words. The model learned to predict the brain activity that reading certain words would trigger. ![]() The idea was to collect a wealth of data the team says is over five times larger than the language data sets typically used in language-related fMRI experiments. “We all like to listen to podcasts, so why not lie in an MRI scanner listening to podcasts?” jokes Alexander Huth, assistant professor of neuroscience and computer science at the University of Texas at Austin, who led the project.ĭuring the study, three participants each listened to 16 hours of different episodes of the same podcasts while in an MRI scanner, plus a couple of TED talks.
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |