Virtual humans have become an essential part of numerous applications, but creating realistic, plausible and emotive digital representations of humans is still challenging, especially for interactive use cases. One example is the “uncanny valley” effect, which suggests that highly photorealistic virtual humans that can still be recognized as artificial feel unappealing.
The aim of NEUROHUM is to enhance the perceptual and emotional realism of virtual humans in immersive media by neurocognitively-guided computer graphics. In an iterative analysis-synthesis loop, neurocognitive features will be extracted in experiments with comprehensive assessments of user behavior, experience, and physiology, mapped onto computational representations and integrated into the generative process for virtual humans. More realistic virtual humans advance immersive media, and representational mappings shed light on the inner workings of neural networks - both artificial and biological.