Climate change exacerbates the occurence of extreme droughts and heatwaves, increasing the frequency and intensity of large wildfires across the globe. Forecasting wildfire danger and uncovering the drivers behind fire events become central for understanding relevant climate‐land surface feedback and aiding wildfire management. In this work, we leverage Deep Learning (DL) to predict the next day's wildfire danger in a fire‐prone part of the Eastern Mediterranean and explainable Artificial Intelligence (xAI) to diagnose model attributions. We implement DL models that capture the temporal and spatio‐temporal context, generalize well for extreme wildfires, and demonstrate improved performance over the traditional Fire Weather Index. Leveraging xAI, we identify the substantial contribution of wetness‐related variables and unveil the temporal focus of the models. The variability of the contribution of the input variables across wildfire events hints into different wildfire mechanisms. The presented methodology paves the way to more robust, accurate, and trustworthy data‐driven anticipation of wildfires.

The modeling of visual attention has gained much interest during the last few years since it allows to efficiently drive complex visual processes to particular areas of images or video frames. Although the literature concerning bottom-up saliency models is vast, we still lack of generic approaches modeling topdown task and context-driven visual attention. Indeed, many top-down models simply modulate the weights associated to low-level descriptors to learn more accurate representations of visual attention than those ones of the generic fusion schemes in bottom-up techniques. In this paper we propose a hierarchical generic probabilistic framework that decomposes the complex process of context-driven visual attention into a mixture of latent subtasks, each of them being in turn modeled as a combination of specific distributions of low-level descriptors. The inclusion of this intermediate level bridges the gap between low-level features and visual attention and enables more comprehensive representations of the later. Our experiments on a dataset in which videos are organized by genre demonstrate that, by learning specific distributions for each video category, we can notably enhance the system performance.Index Terms-Top-down visual attention, hierarchical probabilistic framework, context-aware model.

Presented by miguel ángel fernández torres in partial fulfillment of the requirements for the Degree of Doctor in Multimedia and Communications universidad carlos iii de madrid Advisors: dr. iván gonzález díaz dr. fernando díaz de maría

Wildfire forecasting is of paramount importance for disaster risk reduction and environmental sustainability. We approach daily fire danger prediction as a machine learning task, using historical Earth observation data from the last decade to predict next-day's fire danger. To that end, we collect, pre-process and harmonize an openaccess datacube, featuring a set of covariates that jointly affect the fire occurrence and spread, such as weather conditions, satellite-derived products, topography features and variables related to human activity. We implement a variety of Deep Learning (DL) models to capture the spatial, temporal or spatio-temporal context and compare them against a Random Forest (RF) baseline. We find that either spatial or temporal context is enough to surpass the RF, while a ConvLSTM that exploits the spatio-temporal context performs best with a test Area Under the Receiver Operating Characteristic of 0.926. Our DL-based proof-of-concept provides national-scale daily fire danger maps at a much higher spatial resolution than existing operational solutions. † Institute for Astronomy, Astrophysics, Space Applications and Remote Sensing 1 https://cwfis.cfs.nrcan.gc.ca/background/summary/fdr 2 https://www.afac.com.au/initiative/afdrs 35th Conference on Neural Information Processing Systems (NeurIPS 2021),

Human eye movements while driving reveal that visual attention largely depends on the context in which it occurs. Furthermore, an autonomous vehicle which performs this function would be more reliable if its outputs were understandable. Capsule Networks have been presented as a great opportunity to explore new horizons in the Computer Vision field, due to their capability to structure and relate latent information. In this paper, we present a hierarchical approach for the prediction of eye fixations in autonomous driving scenarios. Context-driven visual attention can be modeled by considering different conditions which, in turn, are represented as combinations of several spatio-temporal features. With the aim of learning these conditions, we have built an encoder-decoder network which merges visual features' information using a global-local definition of capsules. Two types of capsules are distinguished: representational capsules for features and discriminative capsules for conditions. The latter and the use of eye fixations recorded with wearable eye tracking glasses allow the model to learn both to predict contextual conditions and to estimate visual attention, by means of a multi-task loss function. Experiments show how our approach is able to express either frame-level (global) or pixel-wise (local) relationships between features and contextual conditions, allowing for interpretability while maintaining or improving the performance of black-box related systems in the literature. Indeed, our proposal offers an improvement of 29% in terms of information gain with respect to the best performance reported in the literature. INDEX TERMS Top-down visual attention, eye fixation prediction, context-based learning, interpretability, capsule networks, convolutional neural networks, autonomous driving.