English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

We present a comparative evaluation of various techniques for actionrecognition while keeping as many variables as possible controlled. We employtwo categories of Riemannian manifolds: symmetric positive definite matricesand linear subspaces. For both categories we use their corresponding nearestneighbour classifiers, kernels, and recent kernelised sparse representations.We compare against traditional action recognition techniques based on Gaussianmixture models and Fisher vectors (FVs). We evaluate these action recognitiontechniques under ideal conditions, as well as their sensitivity in morechallenging conditions (variations in scale and translation). Despite recentadvancements for handling manifolds, manifold based techniques obtain thelowest performance and their kernel representations are more unstable in thepresence of challenging conditions. The FV approach obtains the highestaccuracy under ideal conditions. Moreover, FV best deals with moderate scaleand translation changes.

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