Search papers, labs, and topics across Lattice.
This paper addresses the problem of extracting topologically correct and watertight boundaries from implicit neural representations of multiple classes, a problem not previously solved. They introduce a 2D boundary extraction algorithm that guarantees topological consistency and allows for setting a minimum level of detail. The algorithm's effectiveness is demonstrated on geological modeling data, highlighting its ability to handle complex topologies.
Finally, a multi-class boundary extraction algorithm that guarantees topological correctness and water-tightness, even with complex topologies.
Surface extraction from implicit neural representations modelling a single class surface is a well-known task. However, there exist no surface extraction methods from an implicit representation of multiple classes that guarantee topological correctness and no holes. In this work, we lay the groundwork by introducing a 2D boundary extraction algorithm for the multi-class case focusing on topological consistency and water-tightness, which also allows for setting minimum detail restraint on the approximation. Finally, we evaluate our algorithm using geological modelling data, showcasing its adaptiveness and ability to honour complex topology.
