EnviroPulse: Providing Feedback About the Expected Affective Valence of the Environment

Interacting with nature is beneficial to a person\&$\#$39;s mental state, but it can sometimes be difficult to find environments that will induce positive affect (e.g., when planning a run). In this paper, we describe EnviroPulse\—a system for automatically determining and communicating the expected affective valence (EAV) of environments to individuals. We describe a prototype that allows this to be used in real-time on a smartphone, but EnviroPulse could easily be incorporated into GPS systems, mapping services, or image-based systems. Our work differs from existing work in affective computing in that, rather than detecting a user\&$\#$39;s affect directly, we automatically determine the EAV of the environment through visual analysis. We present results that suggest our system can determine the EAV of environments. We also introduce real-time affective visual feedback of the calculated EAV of the images, and present results from an informal study suggesting that real-time visual feedback can be used for induction of affect.