Large-scale concept ontology for multimedia

By Milind R. Naphade, Jelena Tešić, Shih-Fu Chang, Winston H. Hsu, L. Kennedy, Alexander G. Hauptmann, Jon Curtis. IEEE Multimedia, 2006

750 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Signal Processing

View Paper: https://www.semanticscholar.org/paper/86dc975f9cbd9a205f8e82fb1db3b61c6b738fa5

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Signal Processing

Citations in the corpus

Evaluating Color Descriptors for Object and Scene Recognition (2010, methodology): “Combinations are compared to the baseline set by the Mediamill Challenge [22] for the 39 LSCOM-Lite categories [35].”
Designing Category-Level Attributes for Discriminative Visual Recognition (2013, background): “The concepts are sometimes manually or automatically organized into a hierarchical structure (ontology) to characterize different levels of semantics [17,26].”
Learning to Separate Object Sounds by Watching Unlabeled Video (2018, background): “Despite a long history of audio-visual video indexing [41,52,63,64,72], objects in video are often analyzed as if they were silent entities in silent environments.”
High-level event recognition in unconstrained videos (2012, background): “One representative work is the LSCOM ontology [95], which defined 1,000+ concepts by carefully considering their utility for video retrieval, feasibility of automatic detection, and observability in actual datasets.”
A Survey on Visual Content-Based Video Indexing and Retrieval (2011, background): “cepts? How should large-scale concept ontology for videos [76] be constructed? How to choose useful generic concept detectors with high retrieval utility? How many useful concepts are”
The Emerging Trends of Multi-Label Learning (2020, methodology): “For example, [212] develops an extreme multi-label classification method for generic video annotation and [234] builds an ontology of 101 concepts from News video as well.”
DeepSentiBank: Visual Sentiment Concept Classification with Deep Convolutional Neural Networks (2014, background): “Concept modeling has been widely studied in multimedia [25, 31], and computer vision (often referred as “attributes”) [9].”
NUS-WIDE: a real-world web image database from National University of Singapore (2009, result): “The 81 concepts are carefully chosen in such a way that: (a) they are consistent with those concepts defined in many other literatures [2][4][ 10 ][15]; (b) they mostly correspond to the frequent tags in Flickr; (c) they have both general concepts such as “animal” and specific concepts such as “dog””
Recognize complex events from static images by fusing deep channels (2015, background): “We also noticed that the concepts in LSCOM are primarily from TV news, and consequently events in personal lives were not thoroughly covered.”
Concept-Based Video Retrieval (2009, background): “3 Detection results [219] for a lexicon [163, 230] of approximately 500 concepts on TRECVID 2006 video data.”
Efficient Object Category Recognition Using Classemes (2010, methodology): “We took the LSCOM CYC ontology dated 2006-06-30 [13], which contains 2832 unique categories.”
Model recommendation: Generating object detectors from few samples (2015, background): “, LSCOM [38]), intermediate representation based on learning classifiers on related tasks (e.”
Content-based visual search learned from social media (2012, methodology): “Given hundreds of single concept detectors trained on well-labeled examples [85], a considerable amount of papers have been published on how to combine the detectors for visual search.”
Domain Transfer SVM for video concept detection (2009, methodology): “semantic concepts are chosen from the LSCOM-lite lexicon [9], a preliminary version of LSCOM, which covers 36 dominant visual concepts present in broadcast news videos, including objects, scenes, locations, people, events and programs.”
Grounding spatial language for video search (2010, background): “[10] describe the LargeScale Concept Ontology for Multimedia (LSCOM), an effort to create a taxonomy of concepts that are automatically extractable from video, that are useful for retrieval, and that cover a wide variety of semantic phenomena.”