multimodalcritiquebearishImage-space learning-based approaches for inverse rendering often suffer from multi-view inconsistencies and lack explicit 3D representation for stable novel view renderingComputer Vision27 Jul 2026http://arxiv.org/abs/2607.02301v1