{"id":"arxiv/2604.06612","url":"https://arxiv.org/abs/2604.06612","tags":["math.NA","cs.LG"],"score":0,"title":"Neural parametric representations for thin-shell shape optimisation","domain":"math.NA","source":"arxiv","authors":["Xiao Xiao","Fehmi Cirak"],"reviews":0,"subject":"math.NA","summary":"Shape optimisation of thin-shell structures requires a flexible, differentiable geometric representation suitable for gradient-based optimisation. We propose a neural parametric geometry representation (NRep) for shells based on a neural network with periodic activation functions. The NRep is defined using a multi-layer perceptron (MLP), which maps the parametric coordinates of mid-surface vertices to their physical coordinates. A structural compliance optimisation problem is posed to optimise the shape of a thin-shell parameterised by the NRep subject to a volume constraint, with the…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-04-08","reactions":{}}