RESUMEN
A key necessity for the safe and autonomous flight of Unmanned Aircraft Systems (UAS) is their reliable perception of the environment, for example, to assess the safety of a landing site. For visual perception, Machine Learning (ML) provides state-of-the-art results in terms of performance, but the path to aviation certification has yet to be determined as current regulation and standard documents are not applicable to ML-based components due to their data-defined properties. However, the European Union Aviation Safety Agency (EASA) published the first usable guidance documents that take ML-specific challenges, such as data management and learning assurance, into account. In this paper, an important concept in this context is addressed, namely the Operational Design Domain (ODD) that defines the limitations under which a given ML-based system is designed to operate and function correctly. We investigated whether synthetic data can be used to complement a real-world training dataset which does not cover the whole ODD of an ML-based system component for visual object detection. The use-case in focus is the detection of humans on the ground to assess the safety of landing sites. Synthetic data are generated using the methods proposed in the EASA documents, namely augmentations, stitching and simulation environments. These data are used to augment a real-world dataset to increase ODD coverage during the training of Faster R-CNN object detection models. Our results give insights into the generation techniques and usefulness of synthetic data in the context of increasing ODD coverage. They indicate that the different types of synthetic images vary in their suitability but that augmentations seem to be particularly promising when there is not enough real-world data to cover the whole ODD. By doing so, our results contribute towards the adoption of ML technology in aviation and the reduction of data requirements for ML perception systems.
RESUMEN
Tuberculosis (TB) still represents a major global health problem affecting over 10 million people worldwide. The gold-standard procedures for TB diagnosis are culture and nucleic acid amplification techniques. In this context, both lipoarabinomannan (LAM) urine test and rapid molecular tests have been major game changers. However, the low sensitivity of the former and the cost and the prohibitive infrastructure requirements to scale-up in endemic regions of the latter, make the improvement of the TB diagnostic landscape a priority. Most forms of life produce extracellular vesicles (EVs), including bacteria despite differences in bacterial cell envelope architecture. We demonstrated that Mycobacterium tuberculosis (Mtb), the causative agent of TB, produces EVs in vitro and in vivo as part of a sophisticated mechanism to manipulate host cellular physiology and to evade the host immune system. In a previous serology study, we showed that the recognition of several mycobacterial extracellular vesicles (MEV) associated proteins could have diagnostic properties. In this study, we pursued to expand the capabilities of MEVs in the context of TB diagnostics by analyzing the composition of MEVs isolated from Mtb cultures submitted to iron starvation and, testing their immunogenicity against a new cohort of serum samples derived from TB+ patients, latent TB-infected (LTBI) patients and healthy donors. We found that despite the stringent condition imposed by iron starvation, Mtb reduces the number of MEV associated proteins relative to iron sufficient conditions. In addition, TB serology revealed three new MEV antigens with specific biomarker capacity. These results suggest the feasibility of developing a point-of-care (POC) device based on selected MEV-associated proteins.
