CVPR Daily - Thursday

21 DAILY CVPR Thursday When we apply this to adults, it’s nice, it’s cool. But I said to my supervisor that we really need to think about people who have difficulty expressing their emotions, like infants, like adults with dementia. We need to focus on people who really need this technology. This is what my PhD research was: applying machine learning to try and understand emotion for newborn babies. Why does this fascinate you more than other areas? Babies cannot express what they feel. There’s a lot of technology right now that tries to simulate human cognitive states, but in the end, we can always get feedback from an adult who can express himself or herself. For babies, they cannot do anything. For everything, they rely on us. When it comes to emotion, it’s very, very important. The entire treatment plan developed for babies depends on our subjective assessment of how they feel. If we cannot understand it, if we are not very accurate about that, we could mess up the entire treatment plan. The baby might be in pain, but we don’t know that. We avoid giving medication. Now, when you look at babies, do you use the things that you learned in order to detect emotions? Right now at NIH, I’m trying to advance healthcare for vulnerable populations. This can be infants, children with autism, people with dementia, and Ghada Zamzmi