5,011 Hits in 5.2 sec

Automatic concept-to-query mapping for web-based concept detector training

Damian Borth, Adrian Ulges, Thomas Michael Breuel
2011 Proceedings of the 19th ACM international conference on Multimedia - MM '11  
Introduction Nowadays, online platforms like YouTube provide massive content for training of visual concept detectors.  ...  Web Demo A web demo is available at: Figure: The query analysis screen of Lookapp. An automatically constructed query can be inspected and if needed modified.  ... 
doi:10.1145/2072298.2072038 dblp:conf/mm/BorthUB11 fatcat:2awzdpz7yvfszjf7vedylh6gly

Automatic Concept Detector Refinement for Large-Scale Video Semantic Annotation

Xueliang Liu, Benoit Huet
2010 2010 IEEE Fourth International Conference on Semantic Computing  
Thanks to this inexpensively obtained set of new training examples the concept detectors can be reinforced and enhanced based on a wider number of unlabeled samples and therefore better adapt to the corpus  ...  Upon training on a small initial set of labeled video shots, the concept detectors are run on the large scale unlabeled corpus in order to identify and select new training samples.  ...  We run the trained concept detectors on those video shots and select automatically new training samples based on the result.  ... 
doi:10.1109/icsc.2010.15 dblp:conf/semco/LiuH10 fatcat:exqt627sdfgtxklzfcxw5tmsa4

Exploiting Multiple Web Resources towards Collecting Positive Training Samples for Visual Concept Learning

Olga Papadopoulou, Vasileios Mezaris
2015 Proceedings of the 5th ACM on International Conference on Multimedia Retrieval - ICMR '15  
The number of images uploaded to the web is enormous and is rapidly increasing. The purpose of our work is to use these for acquiring positive training data for visual concept learning.  ...  We propose an approach which automatically collects positive training samples from the Web by constructing a multitude of text queries and retaining for each query only very few top-ranked images returned  ...  Figure 1 : 1 (a) Typical approach for collecting positive training samples from the Web, for visual concept detector training. (b) Proposed approach.  ... 
doi:10.1145/2671188.2749338 dblp:conf/mir/PapadopoulouM15 fatcat:wiyknuhhfnenvdb4kay5t4vf3m

Keyword-based concept search on consumer photos by web-based kernel function

Po Tun Wu, Yi Hsuan Yang, Kuan Ting Chen, Winston H. Hsu, Tien Hsu Li, Chun Jen Lee
2008 Proceeding of the 16th ACM international conference on Multimedia - MM '08  
We investigate the problem in three folds: (1) the effective concept mapping and selection methods over large-scale concept ontology; (2) the quality and feasibility of the pre-trained concept detectors  ...  Through experiments over large-scale benchmarks, TRECVID and Flickr550, we confirm the effectiveness of concept search in the proposed framework, where the semantic mapping by web-based kernel function  ...  AUTOMATIC CONCEPT MAPPING To map user queries to pre-defined concept lexicons, there are mainly two components, query-concept mapping and concept selection; the former measures the semantic relatedness  ... 
doi:10.1145/1459359.1459451 dblp:conf/mm/WuYCHLL08 fatcat:gbf7c4eytbhbjdqanykyn3cpju

Selecting Relevant Web Trained Concepts for Automated Event Retrieval [article]

Bharat Singh, Xintong Han, Zhe Wu, Vlad I. Morariu, Larry S. Davis
2015 arXiv   pre-print
We propose an event retrieval algorithm that constructs pairs of automatically discovered concepts and then prunes those concepts that are unlikely to be helpful for retrieval.  ...  An alternative to collecting training videos is to train a large semantic concept bank a priori.  ...  The authors would like to thank Yin Cui for providing the initial concepts.  ... 
arXiv:1509.07845v1 fatcat:dg6ykgqagjdm7hpac2xsyrwaom

Selecting Relevant Web Trained Concepts for Automated Event Retrieval

Bharat Singh, Xintong Han, Zhe Wu, Vlad I. Morariu, Larry S. Davis
2015 2015 IEEE International Conference on Computer Vision (ICCV)  
We propose an event retrieval algorithm that constructs pairs of automatically discovered concepts and then prunes those concepts that are unlikely to be helpful for retrieval.  ...  An alternative to collecting training videos is to train a large semantic concept bank a priori.  ...  The authors would like to thank Yin Cui for providing the initial concepts.  ... 
doi:10.1109/iccv.2015.518 dblp:conf/iccv/SinghHWMD15 fatcat:2q52ckj6djefdc25ky3gu26mwe

Tag-based video retrieval by embedding semantic content in a continuous word space

Arnav Agharwal, Rama Kovvuri, Ram Nevatia, Cees G.M. Snoek
2016 2016 IEEE Winter Conference on Applications of Computer Vision (WACV)  
Content-based event retrieval in unconstrained web videos, based on query tags, is a hard problem due to large intra-class variances, and limited vocabulary and accuracy of the video concept detectors,  ...  We present a technique to overcome this gap by using continuous word space representations to explicitly compute query and detector concept similarity.  ...  Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright annotation thereon.  ... 
doi:10.1109/wacv.2016.7477706 dblp:conf/wacv/AgharwalKNS16 fatcat:u6imnmd3mfbuze3itvvowk3zaa

Webly-Supervised Learning of Multimodal Video Detectors

Junwei Liang, Lu Jiang, Alexander Hauptmann
WELL considers multi-modal information including deep learning visual, audio and speech features, to automatically learn accurate video detectors based on the user query.  ...  "Batkid (a kid in batman costume)" or "destroyed buildings", existing methods require manually labeled data to build detectors for searching.  ...  More interestingly, it would be even more valuable to learn detectors from the web videos and apply to personal video collections.  ... 
doi:10.1609/aaai.v31i1.10541 fatcat:h2obwil3hjgb3hm2fmim5vwysm

Utilizing semantic word similarity measures for video retrieval

Yusuf Aytar, Mubarak Shah, Jiebo Luo
2008 2008 IEEE Conference on Computer Vision and Pattern Recognition  
close to the trained detectors.  ...  and visual co-occurrence. (2) A method for retrieval of videos based on their relevance to a user defined text query using the semantic word similarity and visual content of videos.  ...  In fact, we also have access to the trained concept detectors for these concepts from 374 concept detectors set.  ... 
doi:10.1109/cvpr.2008.4587822 dblp:conf/cvpr/AytarSL08 fatcat:vu2wgehocfgvvmje6jfhw5ia7m

Semantic-Based Video Retrieval Survey

Shaimaa Toriah Mohamed Toriah, Atef Zaki Ghalwash, Aliaa A. A. Youssif
2018 Journal of Computer and Communications  
Thus, there is an urgent need to retrieve, organize, and automate videos.  ...  There is a tremendous growth of digital data due to the stunning progress of digital devices which facilitates capturing them. Digital data include image, text, and video.  ...  For each text query, the web search returns the relevant images to query.  ... 
doi:10.4236/jcc.2018.68003 fatcat:qfep2py7ufhwxea7vazpujltja

Semantic context transfer across heterogeneous sources for domain adaptive video search

Yu-Gang Jiang, Chong-Wah Ngo, Shih-Fu Chang
2009 Proceedings of the seventeen ACM international conference on Multimedia - MM '09  
Automatic video search based on semantic concept detectors has recently received significant attention.  ...  Results also show that the FCS is suitable for measuring query-detector similarity, producing better performance to various other popular measures.  ...  We are grateful to Jun Wang for his help in the formulation part and Xiao-Yong Wei for providing baseline results using WordNet.  ... 
doi:10.1145/1631272.1631296 dblp:conf/mm/JiangNC09 fatcat:au3kklsoarb77mnfomme4zjbjy

Visual Learning of Semantic Concepts in Social Multimedia

Damian Borth
2014 Künstliche Intelligenz  
The thesis presents an approach that offers an automatic concept-to-query mapping for training data acquisition from YouTube, where queries are automatically constructed by keyword selection and category  ...  Web Video and Active Relevance Filtering One major challenge in concept detection based on web video is that of how to retrieve proper visual training content from web platforms like YouTube.  ... 
doi:10.1007/s13218-014-0328-x fatcat:hq267u3qnreffcmypn63qn2zmy

VRFP: On-the-fly Video Retrieval using Web Images and Fast Fisher Vector Products [article]

Xintong Han, Bharat Singh, Vlad I. Morariu, Larry S. Davis
2017 arXiv   pre-print
VRFP is a real-time video retrieval framework based on short text input queries, which obtains weakly labeled training images from the web after the query is known.  ...  While a Fisher Vector can be constructed efficiently for a new query, matching against the test set is slow due to its high dimensionality.  ...  ACKNOWLEDGMENT The authors would like to thank Yin Cui, Guangnan Ye, and Yitong Li for providing the AP for each event of their methods.  ... 
arXiv:1512.03384v3 fatcat:buorruy7pbfkfjw3g6npa53kg4

Columbia University/VIREO-CityU/IRIT TRECVID2008 High-Level Feature Extraction and Interactive Video Search

Shih-Fu Chang, Junfeng He, Yu-Gang Jiang, Akira Yanagawa, Eric Zavesky, Elie el Khoury, Chong-Wah Ngo
2008 TREC Video Retrieval Evaluation  
Underlined concepts are unavailable in the closed set; bold concepts are selected by expert and novice.  ...  Table 4: Search topics and semantic concepts proposed in free form and via guided user selection.  ...  Faced with this requirement we can however leverage many of the advances developed for automatic search like utilizing an ontological mapping of terms in query topics, class specific techniques for query  ... 
dblp:conf/trecvid/ChangHJYZKN08 fatcat:v6zwncy3jbe3dm254ajqwrigii

VIREO-TNO @ TRECVID 2014: Multimedia Event Detection and Recounting (MED and MER)

Chong-Wah Ngo, Yi-Jie Lu, Hao Zhang, Ting Yao, Chun Chet Tan, Lei Pang, Maaike de Boer, John G. M. Schavemaker, Klamer Schutte, Wessel Kraaij
2014 TREC Video Retrieval Evaluation  
Late fusion is used to combine classifier scores from multi-SVMs and zero-example system. The final score of a video is adjusted based on OCR matching.  ...  Our 7 runs are summarized below: -MED14Full PS 000Ex: Zero-example system with automatic semantic query generation and OCR matching.  ...  The selection is based on the weighted sum of concept-to-event similarities and concept detectors' responses for all concepts in the semantic query.  ... 
dblp:conf/trecvid/NgoL0YTPBSSK14 fatcat:whrhifh5knhkzfwyxhao6bdkdy
« Previous Showing results 1 — 15 out of 5,011 results