Supervised Learning of Semantic Classes for Image Annotation and Retrieval

By Gustavo Carneiro, Antoni B. Chan, Pedro J. Moreno Mengibar, Nuno Vasconcelos. 2007

999 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Media Technology

How researchers see this paper

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

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

Image-Based Recommendations on Styles and Substitutes (2015, background): “The system shares these characteristics with more common forms of recommender system, but does so on the basis of the appearance of the object, rather than metadata, reviews, or similar.”
Deep Convolutional Ranking for Multilabel Image Annotation (2013, background): “Early works focused on generative model based tagging [1, 3, 26], centred upon the idea of learning a parametric model to perform predictions.”
Beyond Nouns: Exploiting Prepositions and Comparative Adjectives for Learning Visual Classifiers (2008, methodology): “Our work is clearly related to prior work on relating text captions and image features for automatic image annotation [2–4].”
The segmented and annotated IAPR TC-12 benchmark (2010, methodology): “However, due to the lack of a benchmark collection speciﬁcally designed for the requirements of AIA, most methods have been evaluated in small collections of unrealistic images [3–9].”
Multiple-instance ranking: Learning to rank images for image retrieval (2008, methodology): “MI learning has been widely exploited for content-based image retrieval [6, 8, 25], however, they are mainly in the classification setting.”
Multi-label Classification via Feature-aware Implicit Label Space Encoding (2014, methodology): “…it is utilized for tackling numerous real-world applications like multi-label text classification (Katakis et al., 2008), semantic image annotation (Carneiro et al., 2007) and music emotion categorization (Tsoumakas et al., 2008b), etc. Proceedings of the 31 st International Conference on Machine…”
On the Role of Correlation and Abstraction in Cross-Modal Multimedia Retrieval (2014, background): “The existence of a well-known semantic gap, between current image representations and those adopted by humans, severely hampers the performance of unimodal image retrieval systems [2].”
Transductive Multilabel Learning via Label Set Propagation (2013, background): “For example, in image annotation, one image can be tagged with a set of multiple words, such as urban, building, and road, indicating the contents of the image [6], [27].”
Semantic hierarchies for image annotation: A survey (2012, neutral): “[17] based on multiple instance learning.”
Deep Collaborative Embedding for Social Image Understanding (2019, methodology): “Many methods have been proposed to explore images and the associated text information [9], [12], [13], [14], [15], [16], [17], [18], [19], [20].”
Large-scale Multi-label Learning with Missing Labels (2013, background): “Large scale multi-label classiﬁcation is an important learning problem with several applications to real-world problems such as image and video annotation Carneiro et al. (2007); Wang et al. (2009) and query/keyword suggestions Agrawal et al. (2013).”
On the Consistency of Multi-Label Learning (2011, neutral): “For example, in text categorization, a document about national education service may cover several predefined topics, such as government and education, indicating the content in the document (Schapire and Singer, 2000); in bioinformatics, a gene sequence may be relevant to multiple functions, such a”
A new approach to cross-modal multimedia retrieval (2010, background): “In computer vision, substantial effort has been devoted to the problem of image annotation [4, 12, 9, 16, 19, 1].”
Visual Contextual Advertising: Bringing Textual Advertisements to Images (2010, background): “In (Carneiro et al. 2007), image annotation is posed as classification problems where each class is defined by images sharing a common semantic label.”
Evaluating bag-of-visual-words representations in scene classification (2007, methodology): “on multiple instance learning (MIL) [3], and the method proposed by Li and Wang [10].”
A New Baseline for Image Annotation (2008, methodology): “For comparisons on Corel5K, we summarized published results of several approaches, including the most popular topic model (i.e. CorrLDA [4]), as well as MBRM [9] and SML [2], which have shown state-of-the-art performance on Corel5K.”
A review on automatic image annotation techniques (2012, methodology): “Third, for those concept by concept annotation models [26,33], the model for each concept is learned independently and the correlations between concepts are not captured.”