Names and faces in the news

By Tamara L. Berg, Alexander C. Berg, Jaety Edwards, Michael Maire, Ryan White, Yee Whye Teh, Eric Learned-Miller, D.A. Forsyth. 2004

467 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Multimodal Learning

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 Multimodal Learning

Citations in the corpus

End-to-End Learning of Visual Representations From Uncurated Instructional Videos (2019, methodology): “Examples of such methods include learning of person names from image captions or video scripts [3, 10, 35, 37].”
Deep Visual-Semantic Alignments for Generating Image Descriptions (2014, neutral): “A number of approaches have been developed for grounding text in the visual domain [35], [36], [37], [38], [39], [40].”
Data and its (dis)contents: A survey of dataset development and use in machine learning research (2020, background): “the major copyright holders of those images are news wire services, including the Associated Press and Reuters.(142) Other datasets are collected in a studio environment, where images were taken by dataset curators and therefore are copyright holders, which avoids potential copyright issues.”
Using Multiple Segmentations to Discover Objects and their Extent in Image Collections (2006, background): “Examples include discovering main characters [10] and other prominent objects and scenes [23] in movies or mining famous people in collections of news photographs [1].”
An empirical study of context in object detection (2009, background): “Semantic context might indicate the kind of event, activity, or other scene category being depicted [1, 22, 28].”
Training Object Class Detectors from Eye Tracking Data (2014, background): “Another line of work leverages text naturally occurring near an image as a weak annotation for its contents [3,11,19,30], such as on web pages or newspapers.”
DiscLDA: Discriminative Learning for Dimensionality Reduction and Classification (2008, methodology): “This model has been used successfully in a number of applied domains, including information retrieval, vision and bioinformatics [8, 1].”
Attribute and simile classifiers for face verification (2009, background): “’s “Labeled Faces in the Wild” (LFW) benchmark data set [19] and similar data sets [2, 10], 3D alignment is difficult and has not (yet) been demonstrated.”
Harvesting Image Databases from the Web (2007, methodology): “[4] uses text from the Internet, but focused on identifying a specific class rather than general object classes.”
Learning from Partial Labels (2011, methodology): “4 Names and Faces The specific problem of naming faces in images and videos using text sources has been addressed in several works (Satoh et al., 1999; Berg et al., 2004; Gallagher and Chen, 2007; Everingham et al., 2006).”
Learning cross-modality similarity for multinomial data (2011, background): “Several approaches [1, 3, 22, 20, 2] have proposed general probabilistic models to tackle the multi-modal scenario for tasks such as object detection, recognition, and scene understanding.”
Linking People in Videos with "Their" Names Using Coreference Resolution (2014, background): “Another line of work focuses on learning classifiers from images with text [5, 15, 21].”
Unsupervised Joint Alignment of Complex Images (2007, unknown): “In fact, alignment itself has emerged as an important sub-problem in the face recognition literature [18], and a number of systems exist for the detailed alignment of specific categories of objects, such as faces [3, 4, 5, 6, 12, 19, 20].”
FDDB: A benchmark for face detection in unconstrained settings (2010, background): “[2] created a data set that contains images and associated captions extracted from news articles (see Figure 1).”
Shape Matching and Object Recognition (2006, methodology): “The face dataset we use is sampled from the very large dataset used in [6] consisting of news photographs collected from yahoo.”
Exploring photobios (2011, methodology): “Automating this step is also likely possible using face recognition techniques [Berg et al. 2004].”
Randomized Clustering Forests for Image Classification (2008, background): “faces data set [26] is a subset of “Faces in the news”(4) [5] and contains 500 positive and 500 negative pairs of faces.”
Learning realistic human actions from movies (2008, unknown): “Our script-based annotation of human actions is similar in spirit to several recent papers using textual information for automatic image collection from the web [10, 14] and automatic naming of characters in images [1] and videos [4].”
I know what you did last summer: object-level auto-annotation of holiday snaps (2009, background): “In general, auto-annotation is one of the most relevant applications for object recognition methods, and accordingly a large amount of work has been devoted to that problem [2] [4] [5] [11] [12] [19] [23] [24] [25].”