Gaussian Splatting for Real-Time Scene Reconstruction: Methods, Systems, and Trade-offs

Key points

• Since its inception in 2023, its explicit and highly configurable scene representation has been extended toward sparse-view generation, dynamic reconstruction, efficiency and compression, streaming, and simultaneous localization and mapping (SLAM).1
• Although Gaussian Splatting can provide real-time rendering, rendering an existing representation is distinct from constructing, updating, or maintaining the scene itself.1
• Explicitly, we categorize GS scene reconstruction into five overlapping families: single-shot generation, dynamic representations, efficiency and compression, streaming and immersive systems, and Gaussian Splatting SLAM.1
• We distinguish render-only throughput from task-specific scene generation, update, tracking, mapping, and optimization costs, and combine controlled sparse-view benchmarking with native evaluations to identify cross-family quality, latency, memory, and update-cost trade-offs.1

Sources (1)

• [1]Gaussian Splatting for Real-Time Scene Reconstruction: Methods, Systems, and Trade-offs
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 13, 12:20 AM

  3D Gaussian Splatting (3DGS) allows for high-quality real-time novel-view synthesis.

  Since its inception in 2023, its explicit and highly configurable scene representation has been extended toward sparse-view generation, dynamic reconstruction, efficiency and compression, streaming, and simultaneous localization and mapping (SLAM).

Extractive summary: sentences quoted from the sources.