Efficient Multi-View Inverse Rendering Using a Hybrid Differentiable Rendering Method

Authors: Xiangyang Zhu, Yiling Pan, Bailin Deng, Bin Wang

IJCAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We perform quantitative and qualitative evaluation of our method using extensive experiments on both synthetic and real data, and compare it with state-of-the-art methods.
Researcher Affiliation Academia Xiangyang Zhu1,2 , Yiling Pan1,2 , Bailin Deng3 and Bin Wang1,2 1School of Software, Tsinghua University, China 2Beijing National Research Center for Information Science and Technology (BNRist), China 3School of Computer Science and Informatics, Cardiff University, UK
Pseudocode No The paper describes methods and processes in text and equations but does not include structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide an explicit statement or link confirming the open-source release of the authors' own implementation for the described methodology.
Open Datasets Yes We evaluate our method on multiple real-world images from the DTU datasets [Aanæs et al., 2016]... Our synthetic data is created using meshes and textures from [Zhou et al., 2016] and the Internet
Dataset Splits No Concretely, for each object, we render 400 images with colored environmental lighting using graphics engines, 300 for training, and the left 100 for novel view synthesis testing. This specifies train and test splits but no explicit validation split.
Hardware Specification Yes We train and test our method on one NVIDIA RTX 3090 GPU.
Software Dependencies No The paper mentions software like Pytorch3d and Mitsuba 2, but does not provide specific version numbers for them or for PyTorch.
Experiment Setup Yes For geometry optimization, we use the weights (λsil , λlap , λedge , λnormal ) = (1.0, 1.0, 1.0, 0.01) for Eq. 1, and 0.001 for the learning rate. ... For reflectance optimization, we use the weights (λrgb, λreg) = (0.1, 1.0) for Eq. 2, and 0.0001 for the learning rate. To avoid excessive memory consumption in the physically-based rendering module, we set the maximum depth of ray tracing bounce to 3, the downsampling factor of raw images to 4, and the number of sampling per pixel spp to 4 in the iterative optimization phase.