SC-OmniGS: Self-Calibrating Omnidirectional Gaussian Splatting

Authors: Huajian Huang, Yingshu Chen, Longwei Li, Hui Cheng, Tristan Braud, Yajie Zhao, Sai-Kit Yeung

ICLR 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments have demonstrated that our proposed SC-Omni GS is able to recover a high-quality radiance field from noisy camera poses or even no pose prior in challenging scenarios characterized by wide baselines and non-object-centric configurations. The noticeable performance gain in the real-world dataset captured by consumer-grade omnidirectional cameras verifies the effectiveness of our general omnidirectional camera model in reducing the distortion of 360-degree images.
Researcher Affiliation Academia 1 The Hong Kong University of Science and Technology 2 Sun Yat-sen University 3 ICT, University of Southern California
Pseudocode Yes Algorithm 1 illustrates the backpropagation process and the usage of the proposed generic camera model. Algorithm 1: Differentiable Omnidirectional Camera Model
Open Source Code No The paper does not explicitly state that the source code for SC-Omni GS is being released, nor does it provide a direct link to a code repository for the described methodology. It only mentions using official published source codes for baselines.
Open Datasets Yes To verify the efficacy of SC-Omni GS, we conducted extensive experiments using a synthetic dataset Omni Blender (Choi et al., 2023) and a real-world 360Roam dataset (Huang et al., 2022). 360Roam (Huang et al., 2022) provides 360-degree captured images by a consumer-grade 360-degree camera for indoor scenes with multiple rooms, and corresponding initial sparse point clouds from Sf M. All data are under CC BY-NC-SA 4.0 license. Omni Blender (Choi et al., 2023) contains multi-view 360-degree images rendered from Blender synthetic single indoor scenes under MIT License.
Dataset Splits Yes We evaluated SG-Omni GS against several SOTA models on datasets of 360-degree images, including eight real-world multi-room scenes from 360Roam dataset (Huang et al., 2022) each with on average 110 training views and 37 test views, and three synthetic single-room scenes from Omni Blender dataset (Choi et al., 2023) each with 25 training views and 25 test views.
Hardware Specification Yes All methods were run on a desktop computer with an RTX 3090 GPU. On average, for a scene with a Ge Force RTX 3090 GPU, BARF trains for over 2 days, L2G-Ne RF and Cam P for half a day, 3D-GS (six-fold iterations), Omni GS and our SC-Omni GS within 30 minutes.
Software Dependencies No Our SC-Omni GS implementation is built on Pytorch and CUDA. While the software platforms are named, specific version numbers for PyTorch and CUDA are not provided, which are crucial for reproducibility.
Experiment Setup Yes The hyperparameters for 3D Gaussians optimization are set according to the default settings of 3D-GS (Kerbl et al., 2023), with λ = 0.2 and a total of 30,000 optimization iterations. We set the ratio threshold γ to 10. The omnidirectional camera model is shared across all views on individual scene. Moreover, we set the learning rate of the camera model Θ to 1e-4 and activate the angle distortion coefficients D using the Tanh function. For simplicity, we fix ft to 1. The initial learning rates for each camera quaternion q and translation t are set to 0.01, with exponential decay to 1.6e-4 and 6e-3, respectively, in 100 steps per camera. When calibrating from scratch, we increase the initial learning rate of t to 0.1.