Partial model reproduction of "How simple rules determine pedestrian behaviour and crowd disasters" by Moussaid et al. (2010). This version only implements the field of vision rules, no velocity or force rules implemented from the paper. Still, it achieves most of the desired result.
GitHub release here or direct download link from my own domain here.