Weak attributes for large-scale image retrieval

Yu, Felix X.; Ji, Rongrong; Tsai, Ming-Hen; Ye, Guangnan; Chang, Shih-Fu

doi:10.1109/cvpr.2012.6248023

Cited by 70 publications

(70 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Attribute-based representations have recently received much attention because they have been successfully used for image retrieval (Yu et al, 2012), for recognizing objects (Duan et al, 2012;Wang et al, 2009), for describing unknown objects (Farhadi et al, 2009), and even for learning new unseen object models from descriptions (Farhadi et al, 2009;Lampert et al, 2009). Facial attributes have a key role in human-computer interaction applications, image and video retrieval and surveillance.…”

Section: Introductionmentioning

confidence: 99%

Robust gender recognition by exploiting facial attributes dependencies

Bekios-Calfa

Buenaposada

Baumela

2014

Pattern Recognition Letters

View full text Add to dashboard Cite

Estimating human face gender from images is a problem that has been extensively studied because of its relevant applications. Recent works report significant drops in performance for state-of-the-art gender classifiers when evaluated "in the wild," i.e. with uncontrolled demography and environmental conditions. We hypothesize that this is caused by the existence of dependencies among facial demographic attributes that have not been considered when building the classifier. In the paper we study the dependencies among gender, age and pose facial attributes. By considering the relation between gender and pose attributes we also avoid the use of computationally expensive and fragile face alignment procedures. In the experiments we confirm the existence of dependencies among gender, age and pose facial attributes and prove that we can improve the performance and robustness of gender classifiers by exploiting these dependencies.

show abstract

Section: Introductionmentioning

confidence: 99%

Robust gender recognition by exploiting facial attributes dependencies

Bekios-Calfa

Buenaposada

Baumela

2014

Pattern Recognition Letters

View full text Add to dashboard Cite

show abstract

“…Some of these apply to special domains or specific sets of semantic concepts. For example, the space of attributes [19,50,20] is a mid-level semantic representation that has enjoyed substantial popularity in recent years [51][52][53]28,30]. .…”

Section: Related Workmentioning

confidence: 99%

“…Second, because semantic features are, by definition, discriminant for tasks like image categorization, the semantic representation enables the solution of these tasks with low-dimensional classifiers [24,25]. Third, the semantic representation is naturally aligned with recent computer vision interest on contextual modeling [26][27][28][29][30]. This is of importance for tasks such as object recognition, where the detection of contextually related objects has been shown to improve the detection of certain objects of interest [31][32][33], or semantic segmentation, where the coherence of segment semantics can be exploited to achieve more robust segmentations [34][35][36].…”

Section: Introductionmentioning

confidence: 99%

Cross-modal domain adaptation for text-based regularization of image semantics in image retrieval systems

Pereira

Vasconcelos

2014

Computer Vision and Image Understanding

View full text Add to dashboard Cite

a b s t r a c tIn query-by-semantic-example image retrieval, images are ranked by similarity of semantic descriptors. These descriptors are obtained by classifying each image with respect to a pre-defined vocabulary of semantic concepts. In this work, we consider the problem of improving the accuracy of semantic descriptors through cross-modal regularization, based on auxiliary text. A cross-modal regularizer, composed of three steps, is proposed. Training images and text are first mapped to a common semantic space. A regularization operator is then learned for each concept in the semantic vocabulary. This is an operator which maps the semantic descriptors of images labeled with that concept to the descriptors of the associated texts. A convex formulation of the learning problem is introduced, enabling the efficient computation of concept-specific regularization operators. The third step is the selection of the most suitable operator for the image to regularize. This is implemented through a quantization of the semantic space, where a regularization operator is associated with each quantization cell. Overall, the proposed regularizer is a non-linear mapping, implemented as a piecewise linear transformation of the semantic image descriptors to regularize. This transformation is a form of cross-modal domain adaptation. It is shown to achieve better performance than recent proposals in the domain adaptation literature, while requiring much simpler optimization.

show abstract

“…In recent years, attributes have been successfully applied to a variety of computer vision problems including face verification [11], image retrieval [30], action recognition [15], image-totext generation [1]. Category-level attributes are popular not only because they can represent the shared semantic properties of visual classes but because they can leverage information from known categories to enable existing classifiers to generalize to novel categories for which there exists limited training data.…”

Section: Introductionmentioning

confidence: 99%

DaMN – Discriminative and Mutually Nearest: Exploiting Pairwise Category Proximity for Video Action Recognition

Hou¹,

Zamir²,

Sukthankar

et al. 2014

Computer Vision – ECCV 2014

View full text Add to dashboard Cite

Abstract. We propose a method for learning discriminative categorylevel features and demonstrate state-of-the-art results on large-scale action recognition in video. The key observation is that one-vs-rest classifiers, which are ubiquitously employed for this task, face challenges in separating very similar categories (such as running vs. jogging). Our proposed method automatically identifies such pairs of categories using a criterion of mutual pairwise proximity in the (kernelized) feature space, using a category-level similarity matrix where each entry corresponds to the one-vs-one SVM margin for pairs of categories. We then exploit the observation that while splitting such "Siamese Twin" categories may be difficult, separating them from the remaining categories in a two-vs-rest framework is not. This enables us to augment one-vs-rest classifiers with a judicious selection of "two-vs-rest" classifier outputs, formed from such discriminative and mutually nearest (DaMN) pairs. By combining onevs-rest and two-vs-rest features in a principled probabilistic manner, we achieve state-of-the-art results on the UCF101 and HMDB51 datasets. More importantly, the same DaMN features, when treated as a mid-level representation also outperform existing methods in knowledge transfer experiments, both cross-dataset from UCF101 to HMDB51 and to new categories with limited training data (one-shot and few-shot learning). Finally, we study the generality of the proposed approach by applying DaMN to other classification tasks; our experiments show that DaMN outperforms related approaches in direct comparisons, not only on video action recognition but also on their original image dataset tasks.

show abstract

Weak attributes for large-scale image retrieval

Cited by 70 publications

References 20 publications

Robust gender recognition by exploiting facial attributes dependencies

Robust gender recognition by exploiting facial attributes dependencies

Cross-modal domain adaptation for text-based regularization of image semantics in image retrieval systems

DaMN – Discriminative and Mutually Nearest: Exploiting Pairwise Category Proximity for Video Action Recognition

Contact Info

Product

Resources

About