WhittleSearch: Image search with relative attribute feedback

By Adriana Kovashka, Devi Parikh, Kristen Grauman. 2012

350 citations recorded by the source corpus.

Subfields: 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 Computer Vision and Pattern Recognition

Citations in the corpus

Learning Visual Clothing Style with Heterogeneous Dyadic Co-Occurrences (2015, methodology): “To retrieve compatible items, we learn a feature transformation f : Iq → sq from the image space I into the style space S, where compatible items are close together.”
Discovering states and transformations in image collections (2015, background): “As a result, there are many great datasets that cover object [9, 7, 31, 25, 20], attribute [16, 1, 10], material [26], and scene categories [32, 33].”
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”
Deep Learning the City: Quantifying Urban Perception at a Global Scale (2016, background): “The concept of “relative attributes” [37]—ranking object/scene types according to diﬀerent attributes—has been shown to be useful for applications such as image classiﬁcation [38] and guided image search [39].”
VITON: An Image-Based Virtual Try-on Network (2017, background): “, color and textures) of a clothing item [25, 48, 15], since virtual try-on requires preserving the details of a target clothing image as much as possible, including exactly the same style, embroidery, logo, text, etc.”
Towards Transparent Systems: Semantic Characterization of Failure Modes (2014, methodology): “This has been exploited for improved image search by using attributes as keywords [25] or as interactive feedback [24].”
Attribute-Based Classification for Zero-Shot Visual Object Categorization (background): “Attributes have also been suggested as feedback mechanisms to improve image retrieval [51] and categorization [52] .”
Automated Facial Trait Judgment and Election Outcome Prediction: Social Dimensions of Face (2015, background): “Therefore, it is our unique contribution to unite psychological analysis (facial trait perception) and its effects to social construction of leadership (election) within a unified computational framework.”
Interestingness Prediction by Robust Learning to Rank (2014, methodology): “Beyond interesting prediction, many large-scale computer vision problems rely on human intelligence tasks (HIT) using crowdsourcing services, e.g. AMT (Amazon Mechanical Turk) to collect annotations.”
Cross-Domain Image Retrieval with a Dual Attribute-Aware Ranking Network (2015, methodology): “Related to our application domain, Kovashka et al [22] developed a system called “WhittleSearch”, which is able to answer queries such as “Show me shoe images like these, but sportier”.”
Learning the Visual Interpretation of Sentences (2013, background): “Statements about relative attributes can be used to reﬁne search results [18].”
An Attribute-Assisted Reranking Model for Web Image Search (2015, background): “On the other hand, semantic attributes have received tremendous attention recently, where their effectiveness was demonstrated in broad applications, including face verification [6], object recognition [5]–[11], fine-grained visual categorization [16], classification with humans-in-the-loop [22] and”
Text as Neural Operator:Image Manipulation by Text Instruction (2020, background): “tion by text instruction is inspired by cross-modal image retrieval, a fundamental problem in information retrieval for its crucial role in a variety of industrial applications such as product search [23, 54, 11]. In this retrieval setting [47], users search an image database using an input query th”
Hipster Wars: Discovering Elements of Fashion Styles (2014, methodology): “Games for Image Annotation: Large-scale labeling of image data by people is becoming popular in the computer vision community, sometimes using domain experts for very specific labeling tasks [15], and alternatively addressing tasks requiring less specialized expertise or using a combination of human”
COCO Attributes: Attributes for People, Animals, and Objects (2016, methodology): “The Whittlesearch [19] dataset contains 14,658 shoe images with 10 instance-level relative attributes.”
Leveraging Visual Question Answering for Image-Caption Ranking (2016, background): “…improving ﬁne-grained visual recognition, learning models of visual concepts without example images (zero-shot learning [22,23]) and improving human-machine communication where a user can explain the target concept during image search [24,25], or give a classiﬁer an explanation of labels [26,27].”
DeepPermNet: Visual Permutation Learning (2017, methodology): “Due to the expressiveness of visual attributes, researchers have successfully used them for many applications, including image search [23], fine-grained recognition [6] and zero-shot learning [33, 25].”
Discovering Attribute Shades of Meaning with the Crowd (2015, methodology): “…and so trains one classifier (Ferrari and Zisserman, 2007; Kumar et al, 2011; Lampert et al, 2009; Farhadi et al, 2009; Vaquero et al, 2009; Branson et al, 2010; Wang and Mori, 2010; Patterson and Hays, 2012) or one ranking function (Parikh and Grauman, 2011b; Kovashka et al, 2012) per attribute.”