File(s) under permanent embargo
A skeleton-free kinect system for body mass index assessment using deep neural networks
conference contribution
posted on 2017-10-26, 00:00 authored by Darius Nahavandi, Ahmed Abobakr, Hussein Haggag, Mohammed Hossny, Saeid Nahavandi, D Filippidis© 2017 IEEE. In this paper we present a skeleton-free Kinect system to estimate body mass index (BMI) of human bodies. Unlike other systems in the literature, the proposed system does not require a scale to measure the weight. The weight of observed subjects are estimated using body surface area (BSA) regression. The proposed system employs the state-of-the-art deep residual network to extract meaningful features and estimate the BMI scores with a 95% accuracy.