Fast, Dense Feature SDM on an iPhone - Robotics Institute Carnegie Mellon University

Fast, Dense Feature SDM on an iPhone

A. Fagg, S. Sridharan, and S. Lucey
Conference Paper, Proceedings of 12th IEEE International Conference on Automatic Face & Gesture Recognition (FG '17), pp. 95 - 102, May, 2017

Abstract

In this paper, we present our method for enabling dense SDM to run at over 90 FPS on a mobile device. Our contributions are two-fold. Drawing inspiration from the FFT, we propose a Sparse Compositional Regression (SCR) framework, which enables a significant speed up over classical dense regressors. Second, we propose a binary approximation to SIFT features. Binary Approximated SIFT (BASIFT) features, which are a computationally efficient approximation to SIFT, a commonly used feature with SDM. We demonstrate the performance of our algorithm on an iPhone 7, and show that we achieve similar accuracy to SDM.

BibTeX

@conference{Fagg-2017-121039,
author = {A. Fagg and S. Sridharan and S. Lucey},
title = {Fast, Dense Feature SDM on an iPhone},
booktitle = {Proceedings of 12th IEEE International Conference on Automatic Face & Gesture Recognition (FG '17)},
year = {2017},
month = {May},
pages = {95 - 102},
}