Using Computer Vision to Study the Effects of BMI on Online Popularity and Weight-Based Homophily

Increasing prevalence of obesity has disconcerting implications for communities, for nations and, most importantly, for individuals in aspects ranging from quality of life, longevity and health, to social and financial prosperity. Therefore, researchers from a variety of backgrounds study obesity from all angles. In this paper, we use a state-of-the-art computer vision system to predict a person’s body-mass index (BMI) from their social media profile picture and demonstrate the type of analyses this approach enables using data from two culturally diverse settings – the US and Qatar. Using large amounts of Instagram profile pictures, we show that (i) thinner profile pictures have more followers, and that (ii) there is weight-based network homophily in that users with a similar BMI tend to cluster together. To conclude, we also discuss the challenges and limitations related to inferring various user attributes from photos.