On the Geometric Accuracy of Implicit and Primitive-based Representations Derived from View Rendering Constraints
This addresses geometric accuracy requirements for space robotics applications, though it appears incremental in comparing existing methods on a specific dataset.
The paper systematically compares implicit and explicit novel view synthesis methods for 3D object reconstruction in space robotics, finding that appearance embeddings improve photometric fidelity but not geometric accuracy, with convex splatting offering more compact representations than Gaussian splatting.
We present the first systematic comparison of implicit and explicit Novel View Synthesis methods for space-based 3D object reconstruction, evaluating the role of appearance embeddings. While embeddings improve photometric fidelity by modeling lighting variation, we show they do not translate into meaningful gains in geometric accuracy - a critical requirement for space robotics applications. Using the SPEED+ dataset, we compare K-Planes, Gaussian Splatting, and Convex Splatting, and demonstrate that embeddings primarily reduce the number of primitives needed for explicit methods rather than enhancing geometric fidelity. Moreover, convex splatting achieves more compact and clutter-free representations than Gaussian splatting, offering advantages for safety-critical applications such as interaction and collision avoidance. Our findings clarify the limits of appearance embeddings for geometry-centric tasks and highlight trade-offs between reconstruction quality and representation efficiency in space scenarios.