Fast, Dense Feature SDM on an iPhone

TL;DR

This paper introduces a fast, dense SDM method optimized for mobile devices, utilizing a novel SCR framework and binary SIFT features to achieve high speed without sacrificing accuracy.

Contribution

The paper presents a new SCR framework inspired by FFT for speed, and a binary approximation to SIFT features, enabling real-time dense SDM on smartphones.

Findings

Achieves over 90 FPS on an iPhone 7

Maintains similar accuracy to traditional SDM

Demonstrates practical mobile deployment of dense SDM

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.