RainyScape: Unsupervised Rainy Scene Reconstruction using Decoupled Neural Rendering

arXiv:2404.11401v1 Announce Type: new
Abstract: We propose RainyScape, an unsupervised framework for reconstructing clean scenes from a collection of multi-view rainy images. RainyScape consists of two main modules: a neural rendering module and a rain-prediction module that incorporates a predictor network and a learnable latent embedding that captures the rain characteristics of the scene. Specifically, based on the spectral bias property of neural networks, we first optimize the neural rendering pipeline to obtain a low-frequency scene representation. Subsequently, we jointly …

abstract arxiv collection cs.cv embedding framework images modules network neural rendering prediction rain rendering type unsupervised view