Abstract: In the last years, intrinsic image decomposition has gained attention. Most of the state-of-the-art methods are based on the assumption that reflectance changes come along with strong image edges. Recently, user intervention in the recovery problem has proved to be a remarkable source of improvement. In this paper, we propose a novel approach that aims to overcome the shortcomings of pure edge-based methods by introducing strong surface descriptors, such as the color-name descriptor which introduces high-level considerations resembling top-down intervention. We also use a second surface descriptor, termed color-shade, which allows us to include physical considerations derived from the image formation model capturing gradual color surface variations. Both color cues are combined by means of a Markov Random Field. The method is quantitatively tested on the MIT ground truth dataset using different error metrics, achieving state-of-the-art performance.

Abstract: Burst image restoration attempts to effectively utilize the complementary cues appearing in sequential images to produce a high-quality image. Most current methods use all the available images to obtain the reconstructed image. However, using more images for burst restoration is not always the best option regarding reconstruction quality and efficiency, as the images acquired by handheld imaging devices suffer from degradation and misalignment caused by the camera noise and shake. In this paper, we extend the perception-distortion tradeoff theory by introducing multiple-frame information. We propose the area of the unattainable region as a new metric for perception-distortion tradeoff evaluation and comparison. Based on this metric, we analyse the performance of burst restoration from the perspective of the perception-distortion tradeoff under both aligned bursts and misaligned bursts situations. Our analysis reveals the importance of inter-frame alignment for burst restoration and shows that the optimal burst length for the restoration model depends both on the degree of degradation and misalignment.

Abstract: Cross domain image retrieval is a challenging task that implies matching images from one domain to their pairs from another domain. In this paper we focus on fashion image retrieval, which involves matching an image of a fashion item taken by users, to the images of the same item taken in controlled condition, usually by professional photographer. When facing this problem, we have different products
in train and test time, and we use triplet loss to train the network. We stress the importance of proper training of simple architecture, as well as adapting general models to the specific task.

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.

Abstract: Normalization of signals coming from linear sensors is an ubiquitous mechanism of neural adaptation.1 Local interaction between sensors tuned to a particular feature at certain spatial position and neighbor sensors explains a wide range of psychophysical facts including (1) masking of spatial patterns, (2) non-linearities of motion sensors, (3) adaptation of color perception, (4) brightness and chromatic induction, and (5) image quality assessment. Although the above models have formal and qualitative similarities, it does not necessarily mean that the mechanisms involved are pursuing the same statistical goal. For instance, in the case of chromatic mechanisms (disregarding spatial information), different parameters in the normalization give rise to optimal discrimination or adaptation, and different non-linearities may give rise to error minimization or component independence. In the case of spatial sensors (disregarding color information), a number of studies have pointed out the benefits of masking in statistical independence terms. However, such statistical analysis has not been performed for spatio-chromatic induction models where chromatic perception depends on spatial configuration. In this work we investigate whether successful spatio-chromatic induction models,6 increase component independence similarly as previously reported for masking models. Mutual information analysis suggests that seeking an efficient chromatic representation may explain the prevalence of induction effects in spatially simple images. © (2013) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.

Abstract: Image values are the result of a combination of visual information coming from multiple sources. Recovering information from the multiple factors thatproduced an image seems a hard and ill-posed problem. However, it is important to observe that humans develop the ability to interpret images and recognize and isolate specific physical properties of the scene.
Images describing a single physical characteristic of an scene are called intrinsic images. These images would benefit most computer vision tasks which are often affected by the multiple complex effects that are usually found in natural images (e.g. cast shadows, specularities, interreflections...).
In this thesis we analyze the problem of intrinsic image estimation from different perspectives, including the theoretical formulation of the problem, the visual cues that can be used to estimate the intrinsic components and the evaluation mechanisms of the problem.

Abstract: This work is a part of a surveillance system where content- based image retrieval is done in terms of people appearance. Given an image of a person, our work provides an automatic description of his clothing according to the colour, texture and structural composition of its garments. We present a two-stage process composed by image segmentation and a region-based interpretation. We segment an image by modelling it due to an attributed graph and applying a hybrid method that follows a split-and-merge strategy. We propose the interpretation of five cloth combinations that are modelled in a graph structure in terms of region features. The interpretation is viewed as a graph matching with an associated cost between the segmentation and the cloth models. Fi- nally, we have tested the process with a ground-truth of one hundred images.

Abstract: Recognizing human actions in still images is a challenging problem in computer vision due to significant amount of scale, illumination and pose variation. Given the bounding box of a person both at training and test time, the task is to classify the action associated with each bounding box in an image.
Most state-of-the-art methods use the bag-of-words paradigm for action recognition. The bag-of-words framework employing a dense multi-scale grid sampling strategy is the de facto standard for feature detection. This results in a scale invariant image representation where all the features at multiple-scales are binned in a single histogram. We argue that such a scale invariant
strategy is sub-optimal since it ignores the multi-scale information
available with each bounding box of a person.
This paper investigates alternative approaches to scale coding for action recognition in still images. We encode multi-scale information explicitly in three different histograms for small, medium and large scale visual-words. Our first approach exploits multi-scale information with respect to the image size. In our second approach, we encode multi-scale information relative to the size of the bounding box of a person instance. In each approach, the multi-scale histograms are then concatenated into a single representation for action classification. We validate our approaches on the Willow dataset which contains seven action categories: interacting with computer, photography, playing music,
riding bike, riding horse, running and walking. Our results clearly suggest that the proposed scale coding approaches outperform the conventional scale invariant technique. Moreover, we show that our approach obtains promising results compared to more complex state-of-the-art methods.