‘Transforming visual images to cognitive maps’ (360G-Wellcome-200501_Z_16_A)

We will produce a visual environment production suite, "BonVision", that can render virtual reality environments as well as artificial (‘standard’) visual stimuli. BonVision is designed to be used by any lab: it will be easy to use, with one-step installation, online tutorials, clear documentation, technical support and embedded in an online community for development and protocol exchange. The goal is to provide a single framework for experimental design, in any species including humans, that can be published online and enable replication and exchange. Real-time rendering of virtual environments is necessary for next generation understanding of brain function and behaviour, but is prohibitively difficult to implement for non-experts and therefore limited to few laboratories world-wide, each of which develop custom solutions with proprietary or poorly suited software. "BonVision" will combine the rendering precision needed to study vision, the experimental control needed to design and implement repeatable measurements, and elements of game design needed to produce immersive visual environments. No current package combines these elements. The underlying architecture is the influential "Bonsai" framework, providing native integration with experimental hardware for "closed-loop", real-time programming and integration with measurement devices, while being open-source and an open-resource. BonVision would therefore open new research directions in the sensory, perceptual and cognitive sciences. In collaboration with the developers of Bonsai, NeuroGears Ltd, we have already achieved proof-of-principle. We propose to further develop BonVision, improve accessibility, and develop guides and resources to disseminate and educate the wider community of users through workshops, tutorials and technical support.