Jinkai Hu, Chengzhong Yu, Hongli Liu, Lingqi Yan, Yiqian Wu, Xiaogang Jin
Multi-feature RPNN (MRPNN) renders multi-scattered cloud illumination in real time at 1024×1024 resolution, producing results close to the ground truth (a). Thanks to its novel network design, MRPNN supports configurable shading parameters (b), while the prior work only accepts fixed ones. 𝐺 configures Henyey-Greenstein phase function. In addition, our network correctly handles the shadow boundary (c), which was previously a failure case.
We present
Multi-feature Radiance-Predicting Neural Networks (MRPNN), a practical
framework with a lightweight feature fusion neural network for rendering
high-order scattered radiance of participating media in real time. By
reformulating the Radiative Transfer Equation (RTE) through theoretical
examination, we propose transmittance fields, generated at a low cost, as
auxiliary information to help the network better approximate the RTE,
drastically reducing the size of the neural network. The light weight
network efficiently estimates the difficult-to-solve in-scattering term and
allows for configurable shading parameters while improving prediction
accuracy. In addition, we propose a frequency-sensitive stencil design in
order to handle non-cloud shapes, resulting in accurate shadow boundaries.
Results show that our MRPNN is able to synthesize indistinguishable output
compared to the ground truth. Most importantly, MRPNN achieves a speedup of
two orders of magnitude compared to the state-of-the-art, and is able to
render high-quality participating material in real time.