Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
Swimm3R is a novel framework that integrates medium-aware structure-from-motion (SfM) with Underwater Beta Splatting to enhance underwater 3D reconstruction by mitigating the effects of scattering and attenuation. The method employs a feed-forward backbone to incorporate in-air geometric priors and a physics head to accurately estimate underwater imaging parameters and camera poses. Results on the Barbados underwater video dataset demonstrate that Swimm3R significantly improves point cloud recovery and downstream localization metrics, achieving a 1.47 dB increase in PSNR compared to existing methods.
Swimm3R achieves a 1.47 dB improvement in PSNR for underwater 3D reconstruction, addressing the critical challenges of scattering and attenuation.
We propose Swimm3R, a unified framework that combines medium-aware structure-from-motion (SfM) with Underwater Beta Splatting to address scattering- and attenuation-induced failures in underwater 3D reconstruction. Swimm3R distills in-air geometric priors into a feed-forward backbone and uses a physics head to regress underwater image-formation parameters, camera poses, and restored point clouds. Additionally, we introduce Underwater Beta Splatting, which extends Gaussian splatting with Beta primitives and scattering-aware geometric gradients for stable underwater geometry representation. We further establish the Barbados underwater video dataset to demonstrate the effectiveness of our method in challenging underwater environments. On this dataset, Swimm3R robustly recovers underwater scene structure under challenging scattering conditions, yielding coherent seafloor geometry. Using these predicted point clouds, the proposed Underwater Beta Splatting improves average PSNR by $1.47$ dB over WaterSplatting while increasing downstream localization performance by $2.0$ and $2.4$ percentage points in RRA@15 and RTA@15, respectively.