Abstract: Differently from the regular classification task, in ordinal classification there is an order in the classes. As a consequence not all classification errors matter the same: a predicted class close to the groundtruth one is better than predicting a farther away class. To account for this, most previous works employ loss functions based on the absolute difference between the predicted and groundtruth class labels. We argue that there are many cases in ordinal classification where label values are arbitrary (for instance 1. . . C, being C the number of classes) and thus such loss functions may not be the best choice. We instead propose a network architecture that produces not a single class prediction but an ordered vector, or ranking, of all the possible classes from most to least likely. This is thanks to a loss function that compares groundtruth and predicted rankings of these class labels, not the labels themselves. Another advantage of this new formulation is that we can enforce consistency in the predictions, namely, predicted rankings come from some unimodal vector of scores with mode at the groundtruth class. We compare with the state of the art ordinal classification methods, showing
that ours attains equal or better performance, as measured by common ordinal classification metrics, on three benchmark datasets. Furthermore, it is also suitable for a new task on image aesthetics assessment, i.e. most voted score prediction. Finally, we also apply it to building damage assessment from satellite images, providing an analysis of its performance depending on the degree of imbalance of the dataset.

Abstract: In object-class segmentation, one of the most tedious tasks is to manually segment many object examples in order to learn a model of the object category. Yet, there has been little research on reducing the degree of manual annotation for
object-class segmentation. In this work we explore alternative strategies which do not require full manual segmentation of the object in the training set. In particular, we study the use of bounding boxes as a coarser and much cheaper form of segmentation and we perform a comparative study of several Multiple-Instance Learning techniques that allow to obtain a model with this type of weak annotation. We show that some of these methods can be competitive, when used with coarse
segmentations, with methods that require full manual segmentation of the objects. Furthermore, we show how to use active learning combined with this weakly supervised strategy.
As we see, this strategy permits to reduce the amount of annotation and optimize the number of examples that require full manual segmentation in the training set.

Abstract: Image based human detection is of paramount interest due to its potential applications in fields such as advanced driving assistance, surveillance and media analysis. However, even detecting non-occluded standing humans remains a challenge of intensive research. The most promising human detectors rely on classifiers developed in the discriminative paradigm, i.e., trained with labelled samples. However, labeling is a manual intensive step, especially in cases like human detection where it is necessary to provide at least bounding boxes framing the humans for training. To overcome such problem, some authors have proposed the use of a virtual world where the labels of the different objects are obtained automatically. This means that the human models (classifiers) are learnt using the appearance of rendered images, i.e., using realistic computer graphics. Later, these models are used for human detection in images of the real world. The results of this technique are surprisingly good. However, these are not always as good as the classical approach of training and testing with data coming from the same camera, or similar ones. Accordingly, in this paper we address the challenge of using a virtual world for gathering (while playing a videogame) a large amount of automatically labelled samples (virtual humans and background) and then training a classifier that performs equal, in real-world images, than the one obtained by equally training from manually labelled real-world samples. For doing that, we cast the problem as one of domain adaptation. In doing so, we assume that a small amount of manually labelled samples from real-world images is required. To collect these labelled samples we propose a non-standard active learning technique. Therefore, ultimately our human model is learnt by the combination of virtual and real world labelled samples (Fig. 1), which has not been done before. We present quantitative results showing that this approach is valid.

Abstract: Ground-truth data is essential for the objective evaluation of object detection methods in computer vision. Many works claim their method is robust but they support it with experiments which are not quantitatively assessed with regard some ground-truth. This is one of the main obstacles to properly evaluate and compare such methods. One of the main reasons is that creating an extensive and representative ground-truth is very time consuming, specially in the case of video sequences, where thousands of frames have to be labelled. Could such a ground-truth be generated, at least in part, automatically? Though it may seem a contradictory question, we show that this is possible for the case of video sequences recorded from a moving camera. The key idea is transferring existing frame segmentations from a reference sequence into another video sequence recorded at a different time on the same track, possibly under a different ambient lighting. We have carried out experiments on several video sequence pairs and quantitatively assessed the precision of the transformed ground-truth, which prove that our approach is not only feasible but also quite accurate.

Abstract: This paper presents the combined use of gradient and mutual information for infrared and intensity templates matching. We propose to joint: (i) feature matching in a multiresolution context and (ii) information propagation through scale-space representations. Our method consists in combining mutual information with a shape descriptor based on gradient, and propagate them following a coarse-to-fine strategy. The main contributions of this work are: to offer a theoretical formulation towards a multimodal stereo matching; to show that gradient and mutual information can be reinforced while they are propagated between consecutive levels; and to show that they are valid cost functions in multimodal template matchings. Comparisons are presented showing the improvements and viability of the proposed approach.