Assigning a label to each pixel in an image, namely semantic segmentation, has been an important task in computer vision, and has applications in autonomous driving, robotic navigation, localization, and scene understanding. Fully convolutional neural networks have proved to be a successful solution for the task over the years but most of the work being done focuses primarily on accuracy. In this paper, we present a computationally efficient approach to semantic segmentation, while achieving a high mean intersection over union (mIOU), 70.33% on Cityscapes challenge. The network proposed is capable of running realtime on mobile devices. In addition, we make our code and model weights publicly available.

This paper discusses an effective way to fuse multiple tools to make a modern and easy-to-use software to get some key findings about health related topics on online social platforms, with a special focus on the famous Finnish forum, Suomi24. Several natural language processing have been tested and, sometimes, modified in order to accommodate the Finnish language linguistic structure and achieve our tasks of mining the health discussion content. We also explore the ability to monitor and track diseases in Finish open discussion forum. The developed modular system can help clinicians and medical experts to analyze similar forums to identify and track related events that can be correlated with hospital dataset in order to generate new hypotheses that initiate future treatment based approaches.