Image ranking and retrieval based on multi-attribute queries

Siddiquie, Behjat; Feris, Rogério; Davis, Larry S.

doi:10.1109/cvpr.2011.5995329

Cited by 291 publications

(235 citation statements)

References 13 publications

Supporting

Mentioning

234

Contrasting

Order By: Relevance

“…To retrieve images given a query, our model considers the evidence from all object detectors, not just detectors for objects in the query. Similar ideas have been used in [11] for multi-attribute image retrieval. The difference from [11] is that, in addition to object names, our query also considers certain relations between objects.…”

Section: Structured Object Query Modelmentioning

confidence: 94%

“…In contrast, our work considers a more detailed phrase-level representation of queries, and explicitly models the spatial layout of objects. Moreover, we do not assume the existence of reliable detectors as in [11]. Instead, the object locations are treated as latent variables that are implicitly inferred simultaneously with image retrieval.…”

Section: Introductionmentioning

confidence: 99%

“…One limitation of most previous approaches is that they learn a separate ranking function for each query term (single word), and combine the output of single-word queries heuristically for retrieving multi-word queries. Recently, Siddiquie et al [11] built a framework for multi-attribute retrieval which models the co-occurrence between the attributes. In contrast, our work considers a more detailed phrase-level representation of queries, and explicitly models the spatial layout of objects.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Image Retrieval with Structured Object Queries Using Latent Ranking SVM

Lan

Yang

Wang

et al. 2012

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. We consider image retrieval with structured object queriesqueries that specify the objects that should be present in the scene, and their spatial relations. An example of such queries is "car on the road". Existing image retrieval systems typically consider queries consisting of object classes (i.e. keywords). They train a separate classifier for each object class and combine the output heuristically. In contrast, we develop a learning framework to jointly consider object classes and their relations. Our method considers not only the objects in the query ("car" and "road" in the above example), but also related object categories can be useful for retrieval. Since we do not have ground-truth labeling of object bounding boxes on the test image, we represent them as latent variables in our model. Our learning method is an extension of the ranking SVM with latent variables, which we call latent ranking SVM. We demonstrate image retrieval and ranking results on a dataset with more than a hundred of object classes.

show abstract

Section: Structured Object Query Modelmentioning

confidence: 94%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Image Retrieval with Structured Object Queries Using Latent Ranking SVM

Lan

Yang

Wang

et al. 2012

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…Visual attributes are semantic properties of objects (e.g., "fuzzy", "plastic") that serve as a middle ground between low-level features (e.g., color, texture) and high-level categories. Attributes (or "concepts", their counterpart in multimedia retrieval) are known to provide an effective representation for image search [15,10,20,22,4,8,25,7], especially since they permit content-based keyword queries [10,22,7]. While often treated as categorical ("is smiling" vs. "is not smiling"), attributes can more generally be modeled as continuous or relative properties ("is smiling more than X") [16,21].…”

Section: Related Workmentioning

confidence: 99%

Attribute Pivots for Guiding Relevance Feedback in Image Search

Kovashka

Grauman

2013

2013 IEEE International Conference on Computer Vision

View full text Add to dashboard Cite

In interactive image search, a user iteratively refines his results by giving feedback on exemplar images. Active selection methods aim to elicit useful feedback, but traditional approaches suffer from expensive selection criteria and cannot predict informativeness reliably due to the imprecision of relevance feedback. To address these drawbacks, we propose to actively select "pivot" exemplars for which feedback in the form of a visual comparison will most reduce the system's uncertainty. For example, the system might ask, "Is your target image more or less crowded than this image?" Our approach relies on a series of binary search trees in relative attribute space, together with a selection function that predicts the information gain were the user to compare his envisioned target to the next node deeper in a given attribute's tree. It makes interactive search more efficient than existing strategies-both in terms of the system's selection time as well as the user's feedback effort.

show abstract

“…We believe that sharing information between words is extremely important to learn good discriminative representations, and that the use of attributes is one way to achieve this goal. Attributes are semantic properties that can be used to describe images and categories [6], and have recently gained a lot of popularity for image retrieval and classification tasks [6,15,31,32]. Attributes have also shown ability to transfer information in zero-shot learning settings [15] and have been used for feature compression since they usually provide compact descriptors.…”

Section: Supervised Word Representation With Phoc Attributesmentioning

confidence: 99%

Handwritten Word Spotting with Corrected Attributes

Almazán

Gordo

Fornés

et al. 2013

2013 IEEE International Conference on Computer Vision

View full text Add to dashboard Cite

We propose an approach to multi-writer word spotting, where the goal is to find a query word in a dataset comprised of document images. We propose an attributes-based approach that leads to a low-dimensional, fixed-length representation of the word images that is fast to compute and, especially, fast to compare. This approach naturally leads to an unified representation of word images and strings, which seamlessly allows one to indistinctly perform queryby-example, where the query is an image, and query-bystring, where the query is a string. We also propose a calibration scheme to correct the attributes scores based on Canonical Correlation Analysis that greatly improves the results on a challenging dataset. We test our approach on two public datasets showing state-of-the-art results.

show abstract

Image ranking and retrieval based on multi-attribute queries

Cited by 291 publications

References 13 publications

Image Retrieval with Structured Object Queries Using Latent Ranking SVM

Image Retrieval with Structured Object Queries Using Latent Ranking SVM

Attribute Pivots for Guiding Relevance Feedback in Image Search

Handwritten Word Spotting with Corrected Attributes

Contact Info

Product

Resources

About