Search

Semantic segmentation consists of predicting a semantic label for each image pixel. Conventional deep learning models do not take advantage of ordinal relations that might exist in the domain at hand. For example, it is known that the pupil is inside the iris, and the lane markings are inside the road. Such domain knowledge can be employed as constraints to make the model more robust. The current literature on this topic has explored pixel-wise ordinal segmentation methods, which treat each pixel as an independent observation and promote ordinality in its representation. This paper proposes novel spatial ordinal segmentation methods, which take advantage of the structured image space by considering each pixel as an observation dependent on its neighborhood context to also promote ordinal spatial consistency. When evaluated with five biomedical datasets and multiple configurations of autonomous driving datasets, ordinal methods resulted in more ordinally-consistent models, with substantial improvements in ordinal metrics and some increase in the Dice coefficient. It was also shown that the incorporation of ordinal consistency results in models with better generalization abilities., Comment: 12 pages

In many real-world prediction tasks, class labels contain information about the relative order between labels that are not captured by commonly used loss functions such as multicategory cross-entropy. Recently, the preference for unimodal distributions in the output space has been incorporated into models and loss functions to account for such ordering information. However, current approaches rely on heuristics that lack a theoretical foundation. Here, we propose two new approaches to incorporate the preference for unimodal distributions into the predictive model. We analyse the set of unimodal distributions in the probability simplex and establish fundamental properties. We then propose a new architecture that imposes unimodal distributions and a new loss term that relies on the notion of projection in a set to promote unimodality. Experiments show the new architecture achieves top-2 performance, while the proposed new loss term is very competitive while maintaining high unimodality., Comment: 6 pages

Manuscript symbols can be stored, recognized and retrieved from an entropic digital memory that is associative and distributed but yet declarative; memory retrieval is a constructive operation, memory cues to objects not contained in the memory are rejected directly without search, and memory operations can be performed through parallel computations. Manuscript symbols, both letters and numerals, are represented in Associative Memory Registers that have an associated entropy. The memory recognition operation obeys an entropy trade-off between precision and recall, and the entropy level impacts on the quality of the objects recovered through the memory retrieval operation. The present proposal is contrasted in several dimensions with neural networks models of associative memory. We discuss the operational characteristics of the entropic associative memory for retrieving objects with both complete and incomplete information, such as severe occlusions. The experiments reported in this paper add evidence on the potential of this framework for developing practical applications and computational models of natural memory., Comment: 24 pages, 13 figures

The dissolution or growth of a droplet in a host liquid is an important part for processes like chemical extraction, chromatography or emulsification. In this work we look at the dissolution of a pair of vertically aligned droplets immersed in water, both experimentally and with numerical simulations. The liquids used for the droplets are long chain alcohols with a low but finite solubility in water and a significantly lower density than that of the host liquid. Therefore, a solutal plume is formed above of the bottom droplet and natural convection dominates the dissolution process. We monitor the volume of the droplets and the velocity field around them over time. When the liquids of the two droplets are the same, our previously found scaling laws for the Sherwood and Reynolds numbers as functions of the Rayleigh number (Dietrich et al., 2016, J. Fluid Mech.) can be applied to the lower droplet. However, remarkably, when the liquid of the top droplet is different than that of the bottom droplet the volume as function of time becomes non-monotonic, and an oscillatory Marangoni flow at the top droplet is observed. We identify the competition between solutal Marangoni flow and density driven convection as the origin of the oscillation, and numerically model the process., Comment: 29 pages, 8 figures

Interfacial stability is important for many processes involving heat and mass transfer across two immiscible phases. When this transfer takes place in the form of evaporation of a binary solution with one component being more volatile than the other, gradients in surface tension can arise. These gradients can ultimately destabilise the liquid-gas interface. In the present work, we study the evaporation of an ethanol-water solution, for which ethanol has a larger volatility. The solution is contained in a horizontal Hele-Shaw cell which is open from one end to allow for evaporation into air. A Marangoni instability is then triggered at the liquid-air interface. We study the temporal evolution of this instability by observing the effects that it has on the bulk of the liquid. More specifically, the growth of convective cells is visualized with confocal microscopy and the velocity field close to the interface is measured with micro-particle-image-velocimetry. The results of numerical simulations based on quasi 2D equations satisfactorily compare with the experimental observations, even without consideration of evaporative cooling, although this cooling can play an extra role in experiments. Furthermore, a linear stability analysis applied to a simplified version of the quasi 2D equations showed reasonably good agreement with the results from simulations at early times, when the instability has just been triggered and no coarsening has taken place. In particular, we find a critical Marangoni number below which a regime of stability is predicted., Comment: 32 pages, 11 figures, to be published in the Journal of Fluid Mechanics

Service robots need to reason to support people in daily life situations. Reasoning is an expensive resource that should be used on demand whenever the expectations of the robot do not match the situation of the world and the execution of the task is broken down; in such scenarios the robot must perform the common sense daily life inference cycle consisting on diagnosing what happened, deciding what to do about it, and inducing and executing a plan, recurring in such behavior until the service task can be resumed. Here we examine two strategies to implement this cycle: (1) a pipe-line strategy involving abduction, decision-making and planning, which we call deliberative inference and (2) the use of the knowledge and preferences stored in the robot's knowledge-base, which we call conceptual inference. The former involves an explicit definition of a problem space that is explored through heuristic search, and the latter is based on conceptual knowledge including the human user preferences, and its representation requires a non-monotonic knowledge-based system. We compare the strengths and limitations of both approaches. We also describe a service robot conceptual model and architecture capable of supporting the daily life inference cycle during the execution of a robotics service task. The model is centered in the declarative specification and interpretation of robot's communication and task structure. We also show the implementation of this framework in the fully autonomous robot Golem-III. The framework is illustrated with two demonstration scenarios., Comment: 31 pages, 7 figures and 2 appendices