Many successful models for predicting attention in a scene involve three main steps: convolution with set of filters, center-surround mechanism and spatial pooling to construct saliency map. However, integrating information justifying the choice various parameter values remain open problems. In this paper we show that an efficient model color appearance human vision, which contains principled selection parameters as well innate mechanism, can be generalized obtain outperforms...

State-of-the-art object detectors typically use shape information as a low level feature representation to capture the local structure of an object. This paper shows that early fusion and color, is popular in image classification, leads significant drop performance for detection. Moreover, such approaches also yields suboptimal results categories with varying importance color shape. In this we propose attributes explicit Color are compact, computationally efficient, when combined traditional...

Generally the bag-of-words based image representation follows a bottom-up paradigm. The subsequent stages of process: feature detection, description, vocabulary construction and are performed independent intentioned object classes to be detected. In such framework, combining multiple cues as shape color often provides below-expected results.

In this paper we present a parametric model for automatic color naming where each category is modeled as fuzzy set with membership function. The parameters of the functions are estimated in fitting process using data derived from psychophysical experiments. name assignments obtained by agree previous experiments, and therefore high-level color-naming information provided can be useful different computer vision applications use will introduce interesting advantages terms implementation costs,...

In a previous work (X. Otazu, M. Vanrell, & C. A. Párraga, 2008b), we showed how several brightness induction effects can be predicted using simple multiresolution wavelet model (BIWaM). Here present new for chromatic processes (termed Chromatic Induction Wavelet Model or CIWaM), which is also implemented on framework and based similar assumptions related to the spatial frequency contrast surround energy of stimulus. The CIWaM interpreted as very extension BIWaM channels, in our case are...

The segmentation of a single material reflectance is challenging problem due to the considerable variation in image measurements caused by geometry object, shadows, and specularities. combination these effects has been modeled dichromatic reflection model. However, application model real-world images limited unknown acquisition parameters compression artifacts. In this paper, we present robust for shape histogram space. method based on multilocal creaseness analysis which results set ridges...

Finding color representations that are stable to illuminant changes is still an open problem in computer vision. Until now, most approaches have been based on physical constraints or statistical assumptions derived from the scene, whereas very little attention has paid effects selected illuminants final image representation. The novelty of this paper propose perceptual computed corrected images. We define category hypothesis, which weights set feasible according their ability map onto...

Abstract In computer vision, colour naming has been posed as a fuzzy‐set problem where each category is modeled by function that assigns membership value to any given sample. However, the success in automation of this process relies on having an appropriate psychophysical data set for purpose. article we present obtained from colour‐naming experiment. experiment, used scoring method collect judgments adequate fuzzy modeling task. The composed 387 reflectances, their CIELab and Munsell...

The estimation of the illuminant a scene from digital image has been goal large amount research in computer vision. Color constancy algorithms have dealt with this problem by defining different heuristics to select unique solution within feasible set. performance these shown that there is still long way go globally solve as preliminary step In general, evaluation done comparing angular error between estimated chromaticity and canonical illuminant, which highly dependent on dataset. Recently,...

We propose a saliency model termed SIM (saliency by induction mechanisms), which is based on low-level spatiochromatic that has successfully predicted chromatic phenomena. In so doing, we hypothesize the visual mechanisms enhance or suppress image detail are also responsible for making some regions more salient. Moreover, adds geometrical grouplets to complex features such as corners, and relatively simpler edges. Since our been fitted psychophysical data, it largely nonparametric....

In the last years, intrinsic image decomposition has gained attention. Most of state-of-the-art methods are based on assumption that reflectance changes come along with strong edges. Recently, user intervention in recovery problem proved to be a remarkable source improvement. this paper, we propose novel approach aims overcome shortcomings pure edge-based by introducing surface descriptors, such as color-name descriptor which introduces high-level considerations resembling top-down...

Abstract When light is reflected off a surface, there linear relation between the three human photoreceptor responses to incoming and light. Different colored surfaces have different relations. Recently, Philipona O'Regan (2006) showed that when this singular in mathematical sense, then surface perceived as having highly nameable color. Furthermore, white by corresponding precisely one of four psychophysically measured unique hues. However, O'Regan's approach seems unrelated classical...

The characterization of trichromatic cameras is usually done in terms a device-independent color space, such as the CIE 1931 XYZ space. This indeed convenient since it allows testing results against colorimetric measures. We have characterized our camera to represent human cone activation by mapping sensor's (RGB) responses (LMS) through polynomial transformation, which can be “customized” according types scenes we want represent. Here present method test accuracy measures and...

Scene decomposition into its illuminant, shading, and reflectance intrinsic images is an essential step for scene understanding. Collecting image groundtruth data a laborious task. The assumptions on which the ground-truth procedures are based limit their application to simple scenes with single object taken in absence of indirect lighting interreflections. We investigate synthetic research since extraction ground truth straightforward, it allows more realistic situations (e.g, multiple...