Naila Murray, Luca Marchesotti, & Florent Perronnin. (2012). Learning to Rank Images using Semantic and Aesthetic Labels. In 23rd British Machine Vision Conference (110.pp. 1–110.10).
Abstract: Most works on image retrieval from text queries have addressed the problem of retrieving semantically relevant images. However, the ability to assess the aesthetic quality of an image is an increasingly important differentiating factor for search engines. In this work, given a semantic query, we are interested in retrieving images which are semantically relevant and score highly in terms of aesthetics/visual quality. We use large-margin classifiers and rankers to learn statistical models capable of ordering images based on the aesthetic and semantic information. In particular, we compare two families of approaches: while the first one attempts to learn a single ranker which takes into account both semantic and aesthetic information, the second one learns separate semantic and aesthetic models. We carry out a quantitative and qualitative evaluation on a recently-published large-scale dataset and we show that the second family of techniques significantly outperforms the first one.
|
A. Ruiz, Joost Van de Weijer, & Xavier Binefa. (2014). Regularized Multi-Concept MIL for weakly-supervised facial behavior categorization. In 25th British Machine Vision Conference.
Abstract: We address the problem of estimating high-level semantic labels for videos of recorded people by means of analysing their facial expressions. This problem, to which we refer as facial behavior categorization, is a weakly-supervised learning problem where we do not have access to frame-by-frame facial gesture annotations but only weak-labels at the video level are available. Therefore, the goal is to learn a set of discriminative expressions and how they determine the video weak-labels. Facial behavior categorization can be posed as a Multi-Instance-Learning (MIL) problem and we propose a novel MIL method called Regularized Multi-Concept MIL to solve it. In contrast to previous approaches applied in facial behavior analysis, RMC-MIL follows a Multi-Concept assumption which allows different facial expressions (concepts) to contribute differently to the video-label. Moreover, to handle with the high-dimensional nature of facial-descriptors, RMC-MIL uses a discriminative approach to model the concepts and structured sparsity regularization to discard non-informative features. RMC-MIL is posed as a convex-constrained optimization problem where all the parameters are jointly learned using the Projected-Quasi-Newton method. In our experiments, we use two public data-sets to show the advantages of the Regularized Multi-Concept approach and its improvement compared to existing MIL methods. RMC-MIL outperforms state-of-the-art results in the UNBC data-set for pain detection.
|
Bojana Gajic, Ariel Amato, Ramon Baldrich, & Carlo Gatta. (2019). Bag of Negatives for Siamese Architectures. In 30th British Machine Vision Conference.
Abstract: Training a Siamese architecture for re-identification with a large number of identities is a challenging task due to the difficulty of finding relevant negative samples efficiently. In this work we present Bag of Negatives (BoN), a method for accelerated and improved training of Siamese networks that scales well on datasets with a very large number of identities. BoN is an efficient and loss-independent method, able to select a bag of high quality negatives, based on a novel online hashing strategy.
|
Sagnik Das, Hassan Ahmed Sial, Ke Ma, Ramon Baldrich, Maria Vanrell, & Dimitris Samaras. (2020). Intrinsic Decomposition of Document Images In-the-Wild. In 31st British Machine Vision Conference.
Abstract: Automatic document content processing is affected by artifacts caused by the shape
of the paper, non-uniform and diverse color of lighting conditions. Fully-supervised
methods on real data are impossible due to the large amount of data needed. Hence, the
current state of the art deep learning models are trained on fully or partially synthetic images. However, document shadow or shading removal results still suffer because: (a) prior methods rely on uniformity of local color statistics, which limit their application on real-scenarios with complex document shapes and textures and; (b) synthetic or hybrid datasets with non-realistic, simulated lighting conditions are used to train the models. In this paper we tackle these problems with our two main contributions. First, a physically constrained learning-based method that directly estimates document reflectance based on intrinsic image formation which generalizes to challenging illumination conditions. Second, a new dataset that clearly improves previous synthetic ones, by adding a large range of realistic shading and diverse multi-illuminant conditions, uniquely customized to deal with documents in-the-wild. The proposed architecture works in two steps. First, a white balancing module neutralizes the color of the illumination on the input image. Based on the proposed multi-illuminant dataset we achieve a good white-balancing in really difficult conditions. Second, the shading separation module accurately disentangles the shading and paper material in a self-supervised manner where only the synthetic texture is used as a weak training signal (obviating the need for very costly ground truth with disentangled versions of shading and reflectance). The proposed approach leads to significant generalization of document reflectance estimation in real scenes with challenging illumination. We extensively evaluate on the real benchmark datasets available for intrinsic image decomposition and document shadow removal tasks. Our reflectance estimation scheme, when used as a pre-processing step of an OCR pipeline, shows a 21% improvement of character error rate (CER), thus, proving the practical applicability. The data and code will be available at: https://github.com/cvlab-stonybrook/DocIIW.
|
Fahad Shahbaz Khan, Joost Van de Weijer, Sadiq Ali, & Michael Felsberg. (2013). Evaluating the impact of color on texture recognition. In 15th International Conference on Computer Analysis of Images and Patterns (Vol. 8047, pp. 154–162). Springer Berlin Heidelberg.
Abstract: State-of-the-art texture descriptors typically operate on grey scale images while ignoring color information. A common way to obtain a joint color-texture representation is to combine the two visual cues at the pixel level. However, such an approach provides sub-optimal results for texture categorisation task.
In this paper we investigate how to optimally exploit color information for texture recognition. We evaluate a variety of color descriptors, popular in image classification, for texture categorisation. In addition we analyze different fusion approaches to combine color and texture cues. Experiments are conducted on the challenging scenes and 10 class texture datasets. Our experiments clearly suggest that in all cases color names provide the best performance. Late fusion is the best strategy to combine color and texture. By selecting the best color descriptor with optimal fusion strategy provides a gain of 5% to 8% compared to texture alone on scenes and texture datasets.
Keywords: Color; Texture; image representation
|
Maria Vanrell, Naila Murray, Robert Benavente, C. Alejandro Parraga, Xavier Otazu, & Ramon Baldrich. (2011). Perception Based Representations for Computational Colour. In Alain Trémeau S. T. Raimondo Schettini (Ed.), 3rd International Workshop on Computational Color Imaging (Vol. 6626, pp. 16–30). LNCS. Springer-Verlag.
Abstract: The perceived colour of a stimulus is dependent on multiple factors stemming out either from the context of the stimulus or idiosyncrasies of the observer. The complexity involved in combining these multiple effects is the main reason for the gap between classical calibrated colour spaces from colour science and colour representations used in computer vision, where colour is just one more visual cue immersed in a digital image where surfaces, shadows and illuminants interact seemingly out of control. With the aim to advance a few steps towards bridging this gap we present some results on computational representations of colour for computer vision. They have been developed by introducing perceptual considerations derived from the interaction of the colour of a point with its context. We show some techniques to represent the colour of a point influenced by assimilation and contrast effects due to the image surround and we show some results on how colour saliency can be derived in real images. We outline a model for automatic assignment of colour names to image points directly trained on psychophysical data. We show how colour segments can be perceptually grouped in the image by imposing shading coherence in the colour space.
Keywords: colour perception, induction, naming, psychophysical data, saliency, segmentation
|
Joost Van de Weijer, & Fahad Shahbaz Khan. (2013). Fusing Color and Shape for Bag-of-Words Based Object Recognition. In 4th Computational Color Imaging Workshop (Vol. 7786, pp. 25–34). Springer Berlin Heidelberg.
Abstract: In this article we provide an analysis of existing methods for the incorporation of color in bag-of-words based image representations. We propose a list of desired properties on which bases fusing methods can be compared. We discuss existing methods and indicate shortcomings of the two well-known fusing methods, namely early and late fusion. Several recent works have addressed these shortcomings by exploiting top-down information in the bag-of-words pipeline: color attention which is motivated from human vision, and Portmanteau vocabularies which are based on information theoretic compression of product vocabularies. We point out several remaining challenges in cue fusion and provide directions for future research.
Keywords: Object Recognition; color features; bag-of-words; image classification
|
Jordi Roca, Maria Vanrell, & C. Alejandro Parraga. (2012). What is constant in colour constancy? In 6th European Conference on Colour in Graphics, Imaging and Vision (pp. 337–343).
Abstract: Color constancy refers to the ability of the human visual system to stabilize
the color appearance of surfaces under an illuminant change. In this work we studied how the interrelations among nine colors are perceived under illuminant changes, particularly whether they remain stable across 10 different conditions (5 illuminants and 2 backgrounds). To do so we have used a paradigm that measures several colors under an immersive state of adaptation. From our measures we defined a perceptual structure descriptor that is up to 87% stable over all conditions, suggesting that color category features could be used to predict color constancy. This is in agreement with previous results on the stability of border categories [1,2] and with computational color constancy
algorithms [3] for estimating the scene illuminant.
|
Jaime Moreno, Xavier Otazu, & Maria Vanrell. (2010). Local Perceptual Weighting in JPEG2000 for Color Images. In 5th European Conference on Colour in Graphics, Imaging and Vision and 12th International Symposium on Multispectral Colour Science (255–260).
Abstract: The aim of this work is to explain how to apply perceptual concepts to define a perceptual pre-quantizer and to improve JPEG2000 compressor. The approach consists in quantizing wavelet transform coefficients using some of the human visual system behavior properties. Noise is fatal to image compression performance, because it can be both annoying for the observer and consumes excessive bandwidth when the imagery is transmitted. Perceptual pre-quantization reduces unperceivable details and thus improve both visual impression and transmission properties. The comparison between JPEG2000 without and with perceptual pre-quantization shows that the latter is not favorable in PSNR, but the recovered image is more compressed at the same or even better visual quality measured with a weighted PSNR. Perceptual criteria were taken from the CIWaM (Chromatic Induction Wavelet Model).
|
C. Alejandro Parraga, Ramon Baldrich, & Maria Vanrell. (2010). Accurate Mapping of Natural Scenes Radiance to Cone Activation Space: A New Image Dataset. In 5th European Conference on Colour in Graphics, Imaging and Vision and 12th International Symposium on Multispectral Colour Science (50–57).
Abstract: The characterization of trichromatic cameras is usually done in terms of a device-independent color space, such as the CIE 1931 XYZ space. This is indeed convenient since it allows the testing of results against colorimetric measures. We have characterized our camera to represent human cone activation by mapping the camera sensor's (RGB) responses to human (LMS) through a polynomial transformation, which can be “customized” according to the types of scenes we want to represent. Here we present a method to test the accuracy of the camera measures and a study on how the choice of training reflectances for the polynomial may alter the results.
|
Javier Vazquez, G. D. Finlayson, & Maria Vanrell. (2010). A compact singularity function to predict WCS data and unique hues. In 5th European Conference on Colour in Graphics, Imaging and Vision and 12th International Symposium on Multispectral Colour Science (33–38).
Abstract: Understanding how colour is used by the human vision system is a widely studied research field. The field, though quite advanced, still faces important unanswered questions. One of them is the explanation of the unique hues and the assignment of color names. This problem addresses the fact of different perceptual status for different colors.
Recently, Philipona and O'Regan have proposed a biological model that allows to extract the reflection properties of any surface independently of the lighting conditions. These invariant properties are the basis to compute a singularity index that predicts the asymmetries presented in unique hues and basic color categories psychophysical data, therefore is giving a further step in their explanation.
In this paper we build on their formulation and propose a new singularity index. This new formulation equally accounts for the location of the 4 peaks of the World colour survey and has two main advantages. First, it is a simple elegant numerical measure (the Philipona measurement is a rather cumbersome formula). Second, we develop a colour-based explanation for the measure.
|
|
Eduard Vazquez, & Ramon Baldrich. (2008). Colour Image Segmentation in Presence of Shadows. In 4th European Conference on Colour in Graphics, Imaging and Vision Proceedings (383–387).
|
|
Javier Vazquez, Maria Vanrell, & Ramon Baldrich. (2008). Towards a Psychophysical Evaluation of Colour Constancy Algorithms. In 4th European Conference on Colour in Graphics, Imaging and Vision Proceedings (372–377).
|
|
C. Alejandro Parraga, Robert Benavente, Maria Vanrell, & Ramon Baldrich. (2008). Modelling Inter-Colour Regions of Colour Naming Space. In 4th European Conference on Colour in Graphics, Imaging and Vision Proceedings (218–222).
|
Ivet Rafegas, & Maria Vanrell. (2016). Color spaces emerging from deep convolutional networks. In 24th Color and Imaging Conference (pp. 225–230).
Abstract: Award for the best interactive session
Defining color spaces that provide a good encoding of spatio-chromatic properties of color surfaces is an open problem in color science [8, 22]. Related to this, in computer vision the fusion of color with local image features has been studied and evaluated [16]. In human vision research, the cells which are selective to specific color hues along the visual pathway are also a focus of attention [7, 14]. In line with these research aims, in this paper we study how color is encoded in a deep Convolutional Neural Network (CNN) that has been trained on more than one million natural images for object recognition. These convolutional nets achieve impressive performance in computer vision, and rival the representations in human brain. In this paper we explore how color is represented in a CNN architecture that can give some intuition about efficient spatio-chromatic representations. In convolutional layers the activation of a neuron is related to a spatial filter, that combines spatio-chromatic representations. We use an inverted version of it to explore the properties. Using a series of unsupervised methods we classify different type of neurons depending on the color axes they define and we propose an index of color-selectivity of a neuron. We estimate the main color axes that emerge from this trained net and we prove that colorselectivity of neurons decreases from early to deeper layers.
|
Hassan Ahmed Sial, S. Sancho, Ramon Baldrich, Robert Benavente, & Maria Vanrell. (2018). Color-based data augmentation for Reflectance Estimation. In 26th Color Imaging Conference (pp. 284–289).
Abstract: Deep convolutional architectures have shown to be successful frameworks to solve generic computer vision problems. The estimation of intrinsic reflectance from single image is not a solved problem yet. Encoder-Decoder architectures are a perfect approach for pixel-wise reflectance estimation, although it usually suffers from the lack of large datasets. Lack of data can be partially solved with data augmentation, however usual techniques focus on geometric changes which does not help for reflectance estimation. In this paper we propose a color-based data augmentation technique that extends the training data by increasing the variability of chromaticity. Rotation on the red-green blue-yellow plane of an opponent space enable to increase the training set in a coherent and sound way that improves network generalization capability for reflectance estimation. We perform some experiments on the Sintel dataset showing that our color-based augmentation increase performance and overcomes one of the state-of-the-art methods.
|
Graham D. Finlayson, Javier Vazquez, & Fufu Fang. (2021). The Discrete Cosine Maximum Ignorance Assumption. In 29th Color and Imaging Conference (pp. 13–18).
Abstract: the performance of colour correction algorithms are dependent on the reflectance sets used. Sometimes, when the testing reflectance set is changed the ranking of colour correction algorithms also changes. To remove dependence on dataset we can
make assumptions about the set of all possible reflectances. In the Maximum Ignorance with Positivity (MIP) assumption we assume that all reflectances with per wavelength values between 0 and 1 are equally likely. A weakness in the MIP is that it fails to take into account the correlation of reflectance functions between
wavelengths (many of the assumed reflectances are, in reality, not possible).
In this paper, we take the view that the maximum ignorance assumption has merit but, hitherto it has been calculated with respect to the wrong coordinate basis. Here, we propose the Discrete Cosine Maximum Ignorance assumption (DCMI), where
all reflectances that have coordinates between max and min bounds in the Discrete Cosine Basis coordinate system are equally likely.
Here, the correlation between wavelengths is encoded and this results in the set of all plausible reflectances ’looking like’ typical reflectances that occur in nature. This said the DCMI model is also a superset of all measured reflectance sets.
Experiments show that, in colour correction, adopting the DCMI results in similar colour correction performance as using a particular reflectance set.
|
Trevor Canham, Javier Vazquez, D Long, Richard F. Murray, & Michael S Brown. (2021). Noise Prism: A Novel Multispectral Visualization Technique. 31st Color and Imaging Conference, .
Abstract: A novel technique for visualizing multispectral images is proposed. Inspired by how prisms work, our method spreads spectral information over a chromatic noise pattern. This is accomplished by populating the pattern with pixels representing each measurement band at a count proportional to its measured intensity. The method is advantageous because it allows for lightweight encoding and visualization of spectral information
while maintaining the color appearance of the stimulus. A four alternative forced choice (4AFC) experiment was conducted to validate the method’s information-carrying capacity in displaying metameric stimuli of varying colors and spectral basis functions. The scores ranged from 100% to 20% (less than chance given the 4AFC task), with many conditions falling somewhere in between at statistically significant intervals. Using this data, color and texture difference metrics can be evaluated and optimized to predict the legibility of the visualization technique.
|
