The France Telecom Orange Labs (Beijing) Video Semantic Indexing Systems - TRECVID 2010 Notebook Paper
نویسندگان
چکیده
In this paper, we described the latest video semantic indexing systems developed at France Telecom Orange Labs (Beijing). In our previous systems for TRECVID 2009, the features of color, edge, texture and SIFT were used. This year, some new features based on local descriptors were added for performance improvement. Three Full runs (130 concepts) based on later fusion and one Light run (10 concepts) based on early fusion were submitted, among which we compared the results of unsupervised and supervised late fusion. The effect of cross-domain fusion was also investigated. The run of F_A_FTRDBJ-HLF-2_2 achieved our best MAP of 0.075, which was based on a two-step linear weighted fusion of 19 features. In particular, we used a group of unified weights for all concepts. Such a strategy showed good generalization ability on diverse internet video data.
منابع مشابه
The France Telecom Orange Labs ( Beijing ) Video High - level Feature Extraction Systems – TrecVid 2009 Notebook Paper
In this paper, we described the video high-level feature extraction systems developed at France Telecom Orange Labs (Beijing). In our systems, four categories of lowlevel visual features, namely color, edge, texture and SIFT local descriptors, were extracted. Two approaches to fusing the representative capabilities of these visual features were investigated for different runs. Under the setting...
متن کاملFudan University at TRECVID 2010 : Semantic Indexing
In this notebook paper we describe our participation in the NIST TRECVID 2010 evaluation. We took part in semantic indexing task of benchmark this year. For semantic indexing, we submitted 3 automatic runs using only IACC training data: Fudan.TV10.3: this run is based on visual features of keyframes. Fudan.TV10.2: this run is based on visual features of keyframes and object detection. Fudan.TV1...
متن کاملIRIT @ TRECVid 2010 : Hidden Markov Models for Context-aware Late Fusion of Multiple Audio Classifiers
This notebook paper describes the four runs submitted by IRIT at TRECVid 2010 Semantic Indexing task. The four submitted runs can be described and compared as follows: • Run 4 – late fusion (weighted sum) of multiple audio-only classifiers output • Run 3 – context-aware re-rank of run 4 using hidden Markov model • Run 2 – context-aware late fusion of multiple audio classifiers output with hidde...
متن کاملVideos Semantic Indexing using Image Classification
This notebook paper summarizes Team NEC-UIUC’s approaches for TRECVid 2010 Evaluation of Semantic Indexing. Our submissions mainly take advantage of advanced image classification methods using linear coordinate coding (LCC) of local features powered by the distributed computing software Hadoop. For every video shot, we evenly sample key frames and extract dense local features including DHOG and...
متن کاملPaper TRECVid Semantic Indexing of Video: A 6-Year Ret- rospective
Semantic indexing, or assigning semantic tags to video samples, is a key component for content-based access to video documents and collections. The Semantic Indexing task has been run at TRECVid from 2010 to 2015 with the support of NIST and the Quaero project. As with the previous High-Level Feature detection task which ran from 2002 to 2009, the semantic indexing task aims at evaluating metho...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2010