Dr. Splat: Directly Referring 3D Gaussian Splatting via Direct Language Embedding Registration
This addresses the challenge of holistic 3D scene understanding for applications in robotics and AR/VR, though it builds incrementally on 3D Gaussian Splatting techniques.
The paper tackles the problem of open-vocabulary 3D scene understanding by introducing Dr. Splat, which directly associates language embeddings with 3D Gaussians, resulting in significant performance improvements over existing methods in tasks like 3D semantic segmentation and object localization.
We introduce Dr. Splat, a novel approach for open-vocabulary 3D scene understanding leveraging 3D Gaussian Splatting. Unlike existing language-embedded 3DGS methods, which rely on a rendering process, our method directly associates language-aligned CLIP embeddings with 3D Gaussians for holistic 3D scene understanding. The key of our method is a language feature registration technique where CLIP embeddings are assigned to the dominant Gaussians intersected by each pixel-ray. Moreover, we integrate Product Quantization (PQ) trained on general large-scale image data to compactly represent embeddings without per-scene optimization. Experiments demonstrate that our approach significantly outperforms existing approaches in 3D perception benchmarks, such as open-vocabulary 3D semantic segmentation, 3D object localization, and 3D object selection tasks. For video results, please visit : https://drsplat.github.io/