A Comparative Evaluation of Action Recognition Methods via Riemannian Manifolds, Fisher Vectors and GMMs: Ideal and Challenging Conditions

Carvajal, Johanna, Wiliem, Arnold, McCool, Chris and Lovell, Brian C. (2016). A Comparative Evaluation of Action Recognition Methods via Riemannian Manifolds, Fisher Vectors and GMMs: Ideal and Challenging Conditions. In: PAKDD, Workshop on Machine Learning for Sensory Data Analysis (MLSDA), New Zealand, (). 19-22 April 2016.

Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
Author Carvajal, Johanna
Wiliem, Arnold
McCool, Chris
Lovell, Brian C.
Title of paper A Comparative Evaluation of Action Recognition Methods via Riemannian Manifolds, Fisher Vectors and GMMs: Ideal and Challenging Conditions
Conference name PAKDD, Workshop on Machine Learning for Sensory Data Analysis (MLSDA)
Conference location New Zealand
Conference dates 19-22 April 2016
Publication Year 2016
Sub-type Fully published paper
Total pages 14
Collection year 2017
Language eng
Abstract/Summary We present a comparative evaluation of various techniques for action recognition while keeping as many variables as possible controlled. We employ two categories of Riemannian manifolds: symmetric positive definite matrices and linear subspaces. For both categories we use their corresponding nearest neighbour classifiers, kernels, and recent kernelised sparse representations. We compare against traditional action recognition techniques based on Gaussian mixture models and Fisher vectors (FVs). We evaluate these action recognition techniques under ideal conditions, as well as their sensitivity in more challenging conditions (variations in scale and translation). Despite recent advancements for handling manifolds, manifold based techniques obtain the lowest performance and their kernel representations are more unstable in the presence of challenging conditions. The FV approach obtains the highest accuracy under ideal conditions. Moreover, FV best deals with moderate scale and translation changes.
Q-Index Code EX
Q-Index Status Provisional Code
Institutional Status UQ

 
Versions
Version Filter Type
Citation counts: Google Scholar Search Google Scholar
Created: Fri, 27 May 2016, 14:41:28 EST by Arnold Wiliem on behalf of School of Information Technol and Elec Engineering