Learning Visual Attributes

By Vittorio Ferrari, Andrew Zisserman. 2007

485 citations recorded by the source corpus.

Subfields: Representation & Self-Supervised Learning

How researchers see this paper

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

In Representation & Self-Supervised Learning

Citations in the corpus

Describing Textures in the Wild (2013, background): “The study of perceptual attributes of textures has a long history starting from pre-attentive aspects and grouping [16], to coarse high-level attributes [1, 2, 33], to some recent work aimed at discovering such attributes by automatically mining descriptions of images from the Internet [3, 12].”
Transductive Multi-View Zero-Shot Learning (2015, methodology): “Zero-shot learning Multi-class single label zero-shot learning has now been widely studied using attribute-based intermediate semantic layers [3, 6, 10, 11, 19, 25] or data-driven [8, 9, 21, 30] representations.”
Fine-Grained Emotion Strength Transfer, Control and Prediction for Emotional Speech Synthesis (2020, methodology): “Since the explicit annotations of emotion strength are unavailable, we adopt a relative ranking method [16, 17] to represent the local emotion descriptors.”
Cross-Task Weakly Supervised Learning From Instructional Videos (2019, methodology): “This is similar to attributes [12, 13], which have been used in action recognition in the past [23, 33].”
An embarrassingly simple approach to zero-shot learning (2015, background): “This has proved to be particularly useful for images [10, 11, 21], as well as videos [13, 24].”
Evaluation of output embeddings for fine-grained image classification (2014, background): “Attributes [15, 13, 29] provide one such space; they relate different classes through well-known and shared characteristics of objects.”
Latent Embeddings for Zero-Shot Classification (2016, background): “Combining visual information with attributes [7, 8, 11, 17, 20, 28, 27] has also supported fine grained classification.”
Domain Adaptation for Visual Applications: A Comprehensive Survey (2017, background): “An extreme case of few-shot learning is the zero-shot learning [84, 130], where the new task is deduced from previous tasks without the need of any training data available for this task.”
Label-Embedding for Image Classification (2015, neutral): “Attributes have been used for image description [19], [18], [9], caption generation [27], [41], face recognition [29], [51], [10], image retrieval [28], [56], [15], action recognition [32], [69], novelty detection [62] and object classiﬁcation [30], [18], [64], [65], [34], [54], [38].”
Beyond Nouns: Exploiting Prepositions and Comparative Adjectives for Learning Visual Classifiers (2008, background): “al [18] presented an approach to learn visual attributes from a training dataset of positive and negative images using a generative model.”
Multi-Task CNN Model for Attribute Prediction (2015, neutral): “dation, multimedia search and retrieval [5], [13], [14], [20], [30], [35], [36], [38], [43], [49].”
Active Multitask Learning Using Both Latent and Supervised Shared Topics (2014, background): “Computer vision researchers have attempted to overcome this challenge by either gathering large datasets of web images [11, 14, 35] or by formulating new methods that reduce the amount of human supervision required.”
Describing objects by their attributes (2009, background): “Ferrari and Zisserman [9] learn to localize simple color and texture attributes from loose annotations provided by image search.”
A Variational Autoencoder with Deep Embedding Model for Generalized Zero-Shot Learning (2020, background): “…the images of the unseen classes cannot be obtained during the training phase, the semantic information, such as attributes (Farhadi et al. 2009; Ferrari and Zisserman 2008), word vectors (Mikolov et al. 2013) and sentence descriptions (Reed et al. 2016), shared between the classes is usually…”
ConceptMap: Mining Noisy Web Data for Concept Learning (2014, background): “– Unlike most of the recent studies that focus on learning specific types of categories from noisy images downloaded from web (such as objects [17,33], scenes[55], attributes[53,18], and faces [2,43,20] ) we propose a general framework which is applicable to many domains from low level attributes to”
Attribute relation learning for zero-shot classification (2014, background): “Meanwhile, in computer vision community, researchers have suggested that attributes are useful to describe familiar and unfamiliar things [3,7,12,29,31–34], to serve as mid-level features for object classification [9,10,19] and image retrieval [16,17,20,35], and to facilitate zero-shot learning para”
Tabula rasa: Model transfer for object category detection (2011, background): “One solution to this problem is to represent object categories indirectly by their attributes, and to learn to detect attributes that can be applied across multiple categories [10, 12, 13, 7, 19, 29].”
Learning beautiful (and ugly) attributes (2013, background): “These learned attributes have many applications including aesthetic quality prediction, image classification and retrieval.”