The version in the Kent Academic Repository may differ from the final published version. Users are advised to check http://kar.kent.ac.uk for the status of the paper. Users should always cite the published version of record.

No abstract

Benchmarking self-adaptive software systems calls for a new model that takes into account a distinctive characteristic of such systems: alterations over time (i.e., self-achieved modifications or adjustments triggered by changes in the external or internal contexts of the system). Changes are thus a fundamental component of a resilience benchmark, raising an intrinsic research problem: how to identify and select the most realistic and relevant (sequences of) changes to be included in the benchmarking procedure. The problem is that defining a representative changeload would require access to a large amount of field data, which is not available for most systems. In this paper we propose an approach based on risk analysis to tackle this key issue, debating its effectiveness and usability with a simple case study. The procedure, that combines field data with expert knowledge and experimental data, allows moving from the identification of the generic goals of systems in the benchmarking domain to the identification of the most relevant change scenarios (based on probability and impact) that may prevent those systems from achieving their goals.

Image segmentation consists in creating partitions within an image into meaningful areas and objects. It can be used in scene understanding and recognition, in fields like biology, medicine, robotics, satellite imaging, amongst others. In this work we take advantage of the learned model in a deep architecture, by extracting side-outputs at different layers of the network for the task of image segmentation. We study the impact of the amount of side-outputs and evaluate strategies to combine them. A post-processing filtering based on mathematical morphology idempotent functions is also used in order to remove some undesirable noises. Experiments were performed on the publicly available KITTI Road Dataset for image segmentation. Our comparison shows that the use of multiples side outputs can increase the overall performance of the network, making it easier to train and more stable when compared with a single output in the end of the network. Also, for a small number of training epochs (500), we achieved a competitive performance when compared to the best algorithm in KITTI Evaluation Server.

No abstract