Attribute Learning in Large-Scale Datasets

By Olga Russakovsky, Li Fei-Fei. 2010

223 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

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

Citations in the corpus

Conditional Generative Adversarial Nets (2014, methodology): “For image features we pre-train a convolutional model similar to the one from [14] on the full ImageNet dataset with 21,000 labels [16].”
Metric Learning with Adaptive Density Discrimination (2015, methodology): “We attain attribute labels from the Object Attributes dataset (Russakovsky & Fei-Fei, 2010).”
Multi-Task CNN Model for Attribute Prediction (2015, neutral): “dation, multimedia search and retrieval [5], [13], [14], [20], [30], [35], [36], [38], [43], [49].”
Attribute relation learning for zero-shot classification (2014, background): “For the first problem, various representation of transferable knowledge have been proposed, such as distance metrics [23,24], class priors [25,26], discriminating aspects [27,28], and descriptive attributes [3,10,12,29].”
Evaluating knowledge transfer and zero-shot learning in a large-scale setting (2011, background): “The following details and argues for our choices concerning data set, image representation, and learning methodology.”
DOCK: Detecting Objects by Transferring Common-Sense Knowledge (2018, methodology): “While recent works [17,18,39] have demonstrated the exciting potential of transfer-learning on objectcentric datasets like ImageNet, it has not yet been thoroughly explored on more complex scene-centric datasets like MS COCO.”
Fine-Grained Recognition in the Wild: A Multi-task Domain Adaptation Approach (2017, background): “While prior works such as [45, 21, 18] focus on attribute learning, our goal is to improve adaptation using ground truth attribute labels.”
What Else Does Your Biometric Data Reveal? A Survey on Soft Biometrics (2016, background): “Russakovsky and Fei-Fei [38], as well as studies on similarities between faces or objects based on relative attributes by Parikh and Grauman [39] and Zhou et al.”
Learning Deep Representation with Large-Scale Attributes (2015, methodology): “The ImageNet attribute dataset constructed by Olga and Fei-Fei [32] is an important step towards large-scale attribute labels.”
Describing Clothing by Semantic Attributes (2012, methodology): “Similar work has been performed by Russakovsky and Fei-Fei [11], by utilizing attributes and exploring transfer learning for object classification on large-scale datasets.”
Understanding Objects in Detail with Fine-Grained Attributes (2014, background): “Thus, learning attributes from global object or image cues as commonly done [8, 9, 13, 14, 21, 25, 30, 33– 36, 39, 42–44] may in fact pick up contextual evidence more than learning about the attribute per se.”
RepMet: Representative-based metric learning for classification and one-shot object detection (2018, methodology): “In [24], it was shown that if the DML training and clustering of the category instances are alternated during training, then on a variety of challenging fine-grained datasets [14, 20, 19, 26] the DML based classification improves the state-of-the-art even w.”
Deep Learning for Image and Point Cloud Fusion in Autonomous Driving: A Review (2020, methodology): “earches in artiﬁcial intelligence has since regaining momentum. Especially, deep learning based methods have demonstrated great potentials, outperforming the other methods in the ImageNet competition [18] [19] [20]. A. Deep Learning on Image Convolutional Neural Networks (CNNs) are one of the most e”
Crowdsourcing in Computer Vision (2016, unknown): “While originally attributes were modeled as binary categories [Ferrari and Zisserman, 2007, Lampert et al., 2009, Farhadi et al., 2009, Russakovsky and Fei-Fei, 2010, Wang and Mori, 2010], Parikh and Grauman [2011b] proposed to model them as relative properties, where one image has the attribute…”
Unsupervised Learning of Discriminative Attributes and Visual Representations (2016, background): “For our discovered unsupervised attributes, it is not easy to either subjectively name them or compare them to the available human labels (the ImageNet Attribute subset [39] has ambiguous labels in itself) by consensus.”
ConceptMap: Mining Noisy Web Data for Concept Learning (2014, methodology): “Thus, it is encouraging to observe better results in such a large data-set against [47]’s attribute models trained by a sufficiently large training subset.”
Consistent Structural Relation Learning for Zero-Shot Segmentation (2020, background): “The d w -dimensional semantic embeddings could be word representations ( e.g ., word2vec [45] or GloVe embeddings [46]) or class attribute vectors [47].”
On the relationship between visual attributes and convolutional networks (2015, methodology): “A subset of ImageNet (denoted the ImageNet Attribute dataset [16]) comprises 9600 images labeled with 25 binary attributes.”