by Kohler, Tobias and Stommel, Martin
Abstract:
Top athletes are a frequently appearing class of actors in mass media that are analysed alongside with other dominant classes such as heads of state, common people and audience (cf. ). The wide variety of existing depictions of top athletes makes it dicult to derive precise sports related statements and judge their importance. To solve this problem, we perform a cluster analysis of annotated video material from TV annual reviews. This procedure allows us to outline those visual features that dominate the distribution of the data. As a result we are able to prove or disprove, respectively, two hypotheses about the nature of sports related Key Visuals.
Reference:
Kohler, Tobias and Stommel, Martin, "Similarity-Based Clustering of Key Frames of Top Athletes from End-of-Year Reviews on TV", Technical report, TZI Universität Bremen, no. 62, pp. 1–18, 2011.
Bibtex Entry:
@TECHREPORT{Kohler2011,
author = {Kohler, Tobias and Stommel, Martin},
title = {Similarity-Based Clustering of Key Frames of Top Athletes from End-of-Year
Reviews on TV},
institution = {TZI Universit{\"a}t Bremen},
year = {2011},
number = {62},
abstract = {Top athletes are a frequently appearing class of actors in mass media
that are analysed alongside with other dominant classes such as heads
of state, common people and audience (cf. ). The wide variety of
existing depictions of top athletes makes it dicult to derive precise
sports related statements and judge their importance. To solve this
problem, we perform a cluster analysis of annotated video material
from TV annual reviews. This procedure allows us to outline those
visual features that dominate the distribution of the data. As a
result we are able to prove or disprove, respectively, two hypotheses
about the nature of sports related Key Visuals.},
owner = {pmania},
pages = {1--18},
series = {TZI Technical Reports},
timestamp = {2012.11.06}
}