DeepFN: Towards Generalizable Facial Action Unit Recognition with Deep Face Normalization

no code implementations • 3 Mar 2021 • Javier Hernandez, Daniel McDuff, Ognjen, Rudovic, Alberto Fung, Mary Czerwinski

We show that person-independent models yield significantly lower performance (55% average F1 and accuracy across 40 subjects) than person-dependent models (60. 3%), leading to a generalization gap of 5. 3%.

Action Recognition Denoising +2

DeepFaceLIFT: Interpretable Personalized Models for Automatic Estimation of Self-Reported Pain

no code implementations • 9 Aug 2017 • Dianbo Liu, Fengjiao Peng, Andrew Shea, Ognjen, Rudovic, Rosalind Picard

Previous research on automatic pain estimation from facial expressions has focused primarily on "one-size-fits-all" metrics (such as PSPI).

Multi-Task Learning

Deep Structured Learning for Facial Action Unit Intensity Estimation

no code implementations • CVPR 2017 • Robert Walecki, Ognjen, Rudovic, Vladimir Pavlovic, Björn Schuller, Maja Pantic

The goal of this paper is to model these structures and estimate complex feature representations simultaneously by combining conditional random field (CRF) encoded AU dependencies with deep learning.