Abstract: Maximizing both the area coverage and the resolution per target is highly desirable in many applications of computer vision. However, with a limited number of cameras viewing a scene, the two objectives are contradictory. This chapter is dedicated to active vision systems, trying to achieve a trade-off between these two aims and examining the use of high-level reasoning in such scenarios. The chapter starts by introducing different approaches to active cameras configurations. Later, a single active camera system to track a moving object is developed, offering the reader first-hand understanding of the issues involved. Another section discusses practical considerations in building an active vision platform, taking as an example a multi-camera system developed for a European project. The last section of the chapter reflects upon the future trends of using semantic factors to drive smartly coordinated active systems.

Abstract: The Bag of Visual Words (BoVW) is an established representation in computer vision. Taking inspiration from text mining, this representation has proved
to be very effective in many domains. However, in most cases, standard term-weighting schemes are adopted (e.g.,term-frequency or TF-IDF). It remains open the question of whether alternative weighting schemes could boost the
performance of methods based on BoVW. More importantly, it is unknown whether it is possible to automatically learn and determine effective weighting schemes from
scratch. This paper brings some light into both of these unknowns. On the one hand, we report an evaluation of the most common weighting schemes used in text mining, but rarely used in computer vision tasks. Besides, we propose an evolutionary algorithm capable of automatically learning weighting schemes for computer vision problems. We report empirical results of an extensive study in several computer vision problems. Results show the usefulness of the proposed method.

Abstract: The representation (or feature) learning has been an emerging concept in the last years, since it collects a set of techniques that are present in any theoretical or practical methodology referring to artificial intelligence. In computer vision, a very common representation has adopted the form of the well-known Bag of Visual Words. This representation appears implicitly in most approaches where images are described, and is also present in a huge number of areas and domains: image content retrieval, pedestrian detection, human-computer interaction, surveillance, e-health, and social computing, amongst others. The early stages of this dissertation provide an approach for learning visual representations inside evolutionary algorithms, which consists of evolving weighting schemes to improve the BoVW representations for the task of recognizing categories of videos and images. Thus, we demonstrate the applicability of the most common weighting schemes, which are often used in text mining but are less frequently found in computer vision tasks. Beyond learning these visual representations, we provide an approach based on fusion strategies for learning spatiotemporal representations, from multimodal data obtained by depth sensors. Besides, we specially aim at the evolutionary and dynamic modelling, where the temporal factor is present in the nature of the data, such as video sequences of gestures and actions. Indeed, we explore the effects of probabilistic modelling for those approaches based on dynamic programming, so as to handle the temporal deformation and variance amongst video sequences of different categories. Finally, we integrate dynamic programming and generative models into an evolutionary computation framework, with the aim of learning Bags of SubGestures (BoSG) representations and hence to improve the generalization capability of standard gesture recognition approaches. The results obtained in the experimentation demonstrate, first, that evolutionary algorithms are useful for improving the representation of BoVW approaches in several datasets for recognizing categories in still images and video sequences. On the other hand, our experimentation reveals that both, the use of dynamic programming and generative models to align video sequences, and the representations obtained from applying fusion strategies in multimodal data, entail an enhancement on the performance when recognizing some gesture categories. Furthermore, the combination of evolutionary algorithms with models based on dynamic programming and generative approaches results, when aiming at the classification of video categories on large video datasets, in a considerable improvement over standard gesture and action recognition approaches. Finally, we demonstrate the applications of these representations in several domains for human analysis: classification of images where humans may be present, action and gesture recognition for general applications, and in particular for conversational settings within the field of restorative justice

Abstract: Automatic analyzing humans in photographs or videos has great potential applications in computer vision, including medical diagnosis, sports, entertainment, movie editing and surveillance, just to name a few. Body, face and hand are the most studied components of humans. Body has many variabilities in shape and clothing along with high degrees of freedom in pose. Face has many muscles causing many visible deformity, beside variable shape and hair style. Hand is a small object, moving fast and has high degrees of freedom. Adding human characteristics to all aforementioned variabilities makes human analysis quite a challenging task.
In this thesis, we developed human segmentation in different modalities. In a first scenario, we segmented human body and hand in depth images using example-based shape warping. We developed a shape descriptor based on shape context and class probabilities of shape regions to extract nearest neighbors. We then considered rigid affine alignment vs. nonrigid iterative shape warping. In a second scenario, we segmented face in RGB images using convolutional neural networks (CNN). We modeled conditional random field with recurrent neural networks. In our model pair-wise kernels are not fixed and learned during training. We trained the network end-to-end using adversarial networks which improved hair segmentation by a high margin.
We also worked on 3D hand pose estimation in depth images. In a generative approach, we fitted a finger model separately for each finger based on our example-based rigid hand segmentation. We minimized an energy function based on overlapping area, depth discrepancy and finger collisions. We also applied linear models in joint trajectory space to refine occluded joints based on visible joints error and invisible joints trajectory smoothness. In a CNN-based approach, we developed a tree-structure network to train specific features for each finger and fused them for global pose consistency. We also formulated physical and appearance constraints as loss functions.
Finally, we developed a number of applications consisting of human soft biometrics measurement and garment retexturing. We also generated some datasets in this thesis consisting of human segmentation, synthetic hand pose, garment retexturing and Italian gestures.

Abstract: Competitions have become a popular tool in the data science community to solve hard problems, assess the state of the art and spur new research directions. Companies like Kaggle and open source platforms like Codalab connect people with data and a data science problem to those with the skills and means to solve it. Hence, the question arises: What, if anything, could NIPS add to this rich ecosystem?
In 2017, we embarked to find out. We attracted 23 potential competitions, of which we selected five to be NIPS 2017 competitions. Our final selection features competitions advancing the state of the art in other sciences such as “Classifying Clinically Actionable Genetic Mutations” and “Learning to Run”. Others, like “The Conversational Intelligence Challenge” and “Adversarial Attacks and Defences” generated new data sets that we expect to impact the progress in their respective communities for years to come. And “Human-Computer Question Answering Competition” showed us just how far we as a field have come in ability and efficiency since the break-through performance of Watson in Jeopardy. Two additional competitions, DeepArt and AI XPRIZE Milestions, were also associated to the NIPS 2017 competition track, whose results are also presented within this chapter.

Abstract: Videobronchoscopy is a medical imaging technique that allows interactive navigation inside the respiratory pathways and minimal invasive interventions. Tracheal procedures are ordinary interventions that require measurement of the percentage of obstructed pathway for injury (stenosis) assessment. Visual assessment of stenosis in videobronchoscopic sequences requires high expertise of trachea anatomy and is prone to human error. Accurate detection of tracheal rings is the basis for automated estimation of the size of stenosed trachea. Processing of videobronchoscopic images acquired at the operating room is a challenging task due to the wide range of artifacts and acquisition conditions. We present a model of the geometric-appearance of tracheal rings for its detection in videobronchoscopic videos. Experiments on sequences acquired at the operating room, show a performance close to inter-observer variability