Abstract: Recent advances in thermal infrared imaging (LWIR) has allowed its use in applications beyond of the military domain. Nowadays, this new family of sensors is included in different technical and scientific applications. They offer features that facilitate tasks, such as detection of pedestrians, hot spots, differences in temperature, among others, which can significantly improve the performance of a system where the persons are expected to play the principal role. For instance, video surveillance applications, monitoring, and pedestrian detection.
During this dissertation the next question is stated: Could a couple of sensors measuring different bands of the electromagnetic spectrum, as the visible and thermal infrared, be used to extract depth information? Although it is a complex question, we shows that a system of these characteristics is possible as well as their advantages, drawbacks, and potential opportunities.
The matching and fusion of data coming from different sensors, as the emissions registered at visible and infrared bands, represents a special challenge, because it has been showed that theses signals are weak correlated. Therefore, many traditional techniques of image processing and computer vision are not helpful, requiring adjustments for their correct performance in every modality.
In this research an experimental study that compares different cost functions and matching approaches is performed, in order to build a multimodal stereovision system. Furthermore, the common problems in infrared/visible stereo, specially in the outdoor scenes are identified. Our framework summarizes the architecture of a generic stereo algorithm, at different levels: computational, functional, and structural, which can be extended toward high-level fusion (semantic) and high-order (prior).The proposed framework is intended to explore novel multimodal stereo matching approaches, going from sparse to dense representations (both disparity and depth maps). Moreover, context information is added in form of priors and assumptions. Finally, this dissertation shows a promissory way toward the integration of multiple sensors for recovering three-dimensional information.

Abstract: In this paper, we present a page classification application in a banking workflow. The proposed architecture represents administrative document images by merging visual and textual descriptions. The visual description is based on a hierarchical representation of the pixel intensity distribution. The textual description uses latent semantic analysis to represent document content as a mixture of topics. Several off-the-shelf classifiers and different strategies for combining visual and textual cues have been evaluated. A final step uses an n-gram model of the page stream allowing a finer-grained classification of pages. The proposed method has been tested in a real large-scale environment and we report results on a dataset of 70,000 pages.

Abstract: Over the past years, inverse perspective mapping has been successfully applied to several problems in the field of Intelligent Transportation Systems. In brief, the method consists of mapping images to a new coordinate system where perspective effects are removed. The removal of perspective associated effects facilitates road and obstacle detection and also assists in free space estimation. There is, however, a significant limitation in the inverse perspective mapping: the presence of obstacles on the road disrupts the effectiveness of the mapping. The current paper proposes a robust solution based on the use of multimodal sensor fusion. Data from a laser range finder is fused with images from the cameras, so that the mapping is not computed in the regions where obstacles are present. As shown in the results, this considerably improves the effectiveness of the algorithm and reduces computation time when compared with the classical inverse perspective mapping. Furthermore, the proposed approach is also able to cope with several cameras with different lenses or image resolutions, as well as dynamic viewpoints.

Abstract: This paper presents a new model for the task of scene text visual question answering. In this task questions about a given image can only be answered by reading and understanding scene text. Current state of the art models for this task make use of a dual attention mechanism in which one attention module attends to visual features while the other attends to textual features. A possible issue with this is that it makes difficult for the model to reason jointly about both modalities. To fix this problem we propose a new model that is based on an single attention mechanism that attends to multi-modal features conditioned to the question. The output weights of this attention module over a grid of multi-modal spatial features are interpreted as the probability that a certain spatial location of the image contains the answer text to the given question. Our experiments demonstrate competitive performance in two standard datasets with a model that is faster than previous methods at inference time. Furthermore, we also provide a novel analysis of the ST-VQA dataset based on a human performance study. Supplementary material, code, and data is made available through this link.

Abstract: A crucial component of an autonomous vehicle (AV) is the artificial intelligence (AI) is able to drive towards a desired destination. Today, there are different paradigms addressing the development of AI drivers. On the one hand, we find modular pipelines, which divide the driving task into sub-tasks such as perception and maneuver planning and control. On the other hand, we find end-to-end driving approaches that try to learn a direct mapping from input raw sensor data to vehicle control signals. The later are relatively less studied, but are gaining popularity since they are less demanding in terms of sensor data annotation. This paper focuses on end-to-end autonomous driving. So far, most proposals relying on this paradigm assume RGB images as input sensor data. However, AVs will not be equipped only with cameras, but also with active sensors providing accurate depth information (e.g., LiDARs). Accordingly, this paper analyses whether combining RGB and depth modalities, i.e. using RGBD data, produces better end-to-end AI drivers than relying on a single modality. We consider multimodality based on early, mid and late fusion schemes, both in multisensory and single-sensor (monocular depth estimation) settings. Using the CARLA simulator and conditional imitation learning (CIL), we show how, indeed, early fusion multimodality outperforms single-modality.