Abstract
Iterative compilation focused on specialized phase orders (i.e., custom selections of compiler passes and orderings for each program or function) can significantly improve the performance of compiled code. However, phase ordering specialization typically needs to deal with large solution space. A previous approach, evaluated by targeting an x86 CPU, mitigates this issue by first using a training phase on reference codes to produce a small set of high-quality reusable phase orders. This approach then uses these phase orders to compile new codes, without any code analysis. In this paper, we evaluate the viability of using this approach to optimize the GPU execution performance of OpenCL kernels. In addition, we propose and evaluate the use of a heuristic to further reduce the number of evaluated phase orders, by comparing the speedups of the resulting binaries with those of the training phase for each phase order. This information is used to predict which untested phase order is most likely to produce good results (e.g., highest speedup). We performed our measurements using the PolyBench/GPU OpenCL benchmark suite on an NVIDIA Pascal GPU. Without heuristics, we can achieve a geomean execution speedup of 1.64x, using cross-validation, with 5 non-standard phase orders. With the heuristic, we can achieve the same speedup with only 3 non-standard phase orders. This is close to the geomean speedup achieved in our iterative compilation experiments exploring thousands of phase orders. Given the significant reduction in exploration time and other advantages of this approach, we believe that it is suitable for a wide range of compiler users concerned with performance.

Abstract
Providing a price tariff that matches the randomized behavior of residential consumers is one of the major barriers to demand response (DR) implementation. The current trend of DR products provided by aggregators or retailers are not consumer-specific, which poses additional barriers for the engagement of consumers in these programs. In order to address this issue, this paper describes a methodology based on causality inference between DR tariffs and observed residential electricity consumption to estimate consumers' consumption elasticity. It determines the flexibility of each client under the considered DR program and identifies whether the tariffs offered by the DR program affect the consumers' usual consumption or not. The aim of this approach is to aid aggregators and retailers to better tune DR offers to consumer needs and so to enlarge the response rate to their DR programs. We identify a set of critical clients who actively participate in DR events along with the most responsive and least responsive clients for the considered DR program. We find that the percentage of DR consumers who actively participate seem to be much less than expected by retailers, indicating that not all consumers' elasticity is effectively utilized.

Abstract
Breast cancer high survival rate led to an increased interest in the quality of life after treatment, particularly regarding the aesthetic outcome. Currently used aesthetic assessment methods are subjective, which make reproducibility and impartiality impossible. To create an objective method capable of being selected as the gold standard, it is fundamental to detect, in a completely automatic manner, keypoints in photographs of women's torso after being subjected to breast cancer surgeries. This paper proposes a deep and a hybrid model to detect keypoints with high accuracy. Our methods are tested on two datasets, one composed of images with a clean and consistent background and a second one that contains photographs taken under poor lighting and background conditions. The proposed methods represent an improvement in the detection of endpoints, nipples and breast contour for both datasets in terms of average error distance when compared with the current state-of-the-art.

Abstract
An optical fiber anemometer based on a flexible multi-FBG curvature sensor is reported. The probe is comprised of a structured polymer shell with embedded single-mode fibers with written fiber Bragg gratings. When the sensor is bent, the different spectral shift of the Bragg wavelengths allows the determination of the mechanical stimulus. Moreover, the probe was also used as a cantilever sensor for assessing the airflow speed in a wind tunnel. The sensor presented sensitivities of 0.8 nm/m(-1) and 1.05 pm/(m/s) for curvature and square speed measurements, respectively, and the sensing characteristics can be improved by simply changing the material and the geometry of the bulk polymer shell, providing a versatile and feasible probe for the mechanical and flow measurements.

Abstract
In this work, the fabrication of optical waveguides embedded in fused silica (Suprasil1) and boro-aluminosilicate glass (Eagle2000) is demonstrated with femtosecond laser direct writing, as well as their suitability to be brought to the surface, through wet etching, for enhanced evanescent coupling with the external dielectric medium. Fused silica demonstrated to be inappropriate in this particular application, as the guiding region is at the bottom of the induced modification, creating a barrier between the guided mode and the substrate’s boundary. Furthermore, the existence of nanogratings meant that, upon contact of the top of the induced modification with the substrate’s boundary, the waveguide is quickly etched. Eagle2000 demonstrated to be superior to fused silica due to its characteristic modification cross-section and absence of nanogratings, which allowed the placement of the guiding region as close to the substrate’s surface as required. However, surface roughness arising from the creation of insoluble products in the HF solution was found. The addition of HCl to dissolve these products was implemented.

Abstract
The possibility to engage in autonomous flight through geolocation-based missions turns Unmanned Aerial Vehicles (UAV) into valuable tools that save time and resources in services like deliveries and surveillance. Amazon is already developing a drop-by delivery service, but there are limitations regarding the client’s id, that can be analyzed in three phases: the approach to the potential receiver, the authorization through the client id and the delivery itself. This work shows a solution for the first of these phases. Firstly, the receiver identifies the GPS coordinates where he wants to receive the package. The UAV flights to that place and tries to locate the receiver on the arrival through Computer Vision (CV) techniques, more precisely Deep Neural Networks (DNN), to continue to the next phase, the identification. After the proposal of the system’s architecture and the prototype’s implementation, a test scenario to analyze the feasibility of the proposed techniques was created. The results were quite good considering a system to look for one person in a limited area defined by the destination coordinates, confirming the detection of one person with an up to 92% accuracy from a 10 m height and 5 m horizontal distance in low resolution images. © Springer Nature Switzerland AG 2020.