Curvature-Regularized Variational Autoencoder for 3D Scene Reconstruction from Sparse Depth
This addresses geometric errors in sparse reconstruction for autonomous vehicles and robots, though it appears incremental as it builds on existing variational autoencoder frameworks.
The paper tackles 3D scene reconstruction from sparse depth measurements (only 5% of needed data) by proposing a curvature regularization approach using a discrete Laplacian operator, achieving 18.1% better reconstruction accuracy than standard variational autoencoders.
When depth sensors provide only 5% of needed measurements, reconstructing complete 3D scenes becomes difficult. Autonomous vehicles and robots cannot tolerate the geometric errors that sparse reconstruction introduces. We propose curvature regularization through a discrete Laplacian operator, achieving 18.1% better reconstruction accuracy than standard variational autoencoders. Our contribution challenges an implicit assumption in geometric deep learning: that combining multiple geometric constraints improves performance. A single well-designed regularization term not only matches but exceeds the effectiveness of complex multi-term formulations. The discrete Laplacian offers stable gradients and noise suppression with just 15% training overhead and zero inference cost. Code and models are available at https://github.com/Maryousefi/GeoVAE-3D.