Neural networks are currently suggested to be implemented in several different driving functions of autonomous vehicles. While showing promising results the drawback lies in the difficulty of safety verification and ensuring operation as intended. The aim of this paper is to increase safety when using neural networks, by proposing a monitoring framework based on novelty estimation of incoming driving data. The idea is to use unsupervised instance discrimination to learn a similarity measure across ego-vehicle camera images. By estimating a von Mises-Fisher distribution of expected ego-camera images they can be compared with unexpected novel images. A novelty measurement is inferred through the likelihood of test frames belonging to the expected distribution. The suggested method provides competitive results to several other novelty or anomaly detection algorithms on the CIFAR-10 and CIFAR-100 datasets.It also shows promising results on real world driving scenarios by distinguishing novel driving scenes from the training data of BDD100k. Applied on the identical training-test data split, the method is also able to predict the performance profile of a segmentation network. Finally, examples are provided on how this method can be extended to find novel segments in images.
The quantity of data generated within healthcare is increasing exponentially. Following this development, the interest of using data driven methodologies such as machine learning is on a steady rise. However, the quality of the data also needs to be considered, since information generated for human interpretation may not be optimal for quantitative computer-based analysis. This work investigates dimensions of data quality for the purpose of artificial intelligence applications in healthcare. Particularly, ECG is studied which traditionally rely on analog prints for initial examination. A digitalization process for ECG is implemented, together with a machine learning model for heart failure prediction, to quantitatively compare results based on data quality. The digital time series data provide a significant accuracy increase, compared to scans of analog plots.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.