Informedia at TRECVID2014: MED and MER, Semantic Indexing, Surveillance Event Detection
Abstract
We report on our results in the TRECVID 2011 Multimedia Event Detection (MED) and Semantic Indexing (SIN) tasks. Generally, both of these tasks consist of three main steps: extracting features, training detectors and fusing. In the feature extraction part, we extracted many low-level features, high-level features and text features. We used the Spatial-Pyramid Matching technique to represent the low-level visual local features, such as SIFT and MoSIFT, which describe the location information of feature points. In the detector training part, besides the traditional SVM, we proposed a Sequential Boosting SVM classifier to deal with the large-scale unbalanced classification problem. In the fusion part, to take the advantages from different features, we tried three different fusion methods: early fusion, late fusion and double fusion. Double fusion is a combination of early fusion and late fusion. The experimental results demonstrated that double fusion is consistently better than or at least comparable to early fusion and late fusion.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 10, 2014
- Accession Number
- AD1024376
Entities
People
- Alexander Hauptmann
- Anil Armagan
- Anurag Kumar
- Bhiksha Raj
- Chuang Gan
- Deyu Meng
- Florian Metze
- Huan Li
- Lara Martin
- Lu Jiang
- Ming C. Lin
- Nikolas Wolfe
- Pinar D. Sahin
- Richard M. Stern
- Rita Singh
- Shicheng Xu
- Shiguang Shan
- Shoou-i Yu
- Susanne Burger
- Teruko Mitamura
- Xiaojun Chang
- Xingzhong Du
- Xuanching Li
- Yajie Miao
- Yang Cai
- Yi Yang
- Yicheng Zhao
- Zexi Mao
- Zhenzhong Lan
- Zhigang Ma
- Zhongwen Xu
Organizations
- Carnegie Mellon University