This thesis will focus on the automatic characterisation of an audio recording from urban outdoor spaces with regards to its soundscape quality. Soundscape quality is a qualitative metric that tells us how does a sound makes us feel and it is strongly related to the public's wellness. In an extreme binary case, sounds can be classified either as unbearable or optimal. Obviously, intermediate situations can also be considered. The goal of this thesis is to explore deep learning and transfer learning techniques to automatically assess the soundscape quality. Material: http://users.iit.demokritos.gr/~tyianak/soundscape/