Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

4D Effect Classification by Encoding CNN Features

Abstract: 

4D effects are physical effects simulated in sync with videos, movies, and games to augment the events occurring in a story or a virtual world. Types of 4D effects commonly used for the immersive media may include seat motion, vibration, flash, wind, water, scent, thunderstorm, snow, and fog. Currently, the recognition of physical effects from a video is mainly conducted by human experts. Although 4D effects are promising in giving immersive experience and entertainment, this manual production has been the main obstacle to faster and wider application of 4D effects. In this paper, we utilize pre-trained models of Convolutional Neural Networks (CNNs) to extract local visual features and propose a new representation method that combines extracted features into video level features. Classification tasks are conducted by employing Support Vector Machine (SVM). Comprehensive experiments are performed to investigate different architecture of CNNs and different type of features for 4D effect classification task and compare baseline average pooling method with our proposed video level representation. Our framework outperforms the baseline up to 2-3% in terms of mean average precision (mAP).

up
0 users have voted:

Paper Details

Authors:
Thomhert S. Siadari, Mikyong Han, Hyunjin Yoon
Submitted On:
14 September 2017 - 3:13am
Short Link:
Type:
Presentation Slides
Event:
Presenter's Name:
Thomhert S. Siadari
Paper Code:
WP-L6.6
Document Year:
2017
Cite

Document Files

presentation slides

(24 downloads)

Subscribe

[1] Thomhert S. Siadari, Mikyong Han, Hyunjin Yoon, "4D Effect Classification by Encoding CNN Features", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1892. Accessed: Nov. 23, 2017.
@article{1892-17,
url = {http://sigport.org/1892},
author = {Thomhert S. Siadari; Mikyong Han; Hyunjin Yoon },
publisher = {IEEE SigPort},
title = {4D Effect Classification by Encoding CNN Features},
year = {2017} }
TY - EJOUR
T1 - 4D Effect Classification by Encoding CNN Features
AU - Thomhert S. Siadari; Mikyong Han; Hyunjin Yoon
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1892
ER -
Thomhert S. Siadari, Mikyong Han, Hyunjin Yoon. (2017). 4D Effect Classification by Encoding CNN Features. IEEE SigPort. http://sigport.org/1892
Thomhert S. Siadari, Mikyong Han, Hyunjin Yoon, 2017. 4D Effect Classification by Encoding CNN Features. Available at: http://sigport.org/1892.
Thomhert S. Siadari, Mikyong Han, Hyunjin Yoon. (2017). "4D Effect Classification by Encoding CNN Features." Web.
1. Thomhert S. Siadari, Mikyong Han, Hyunjin Yoon. 4D Effect Classification by Encoding CNN Features [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1892