AsySplat: Efficient Asymmetric 3D Gaussian Splatting for Long-Sequence Scene Modeling

  • 类型:arxiv
  • 标识:2607.10995
  • 链接:https://arxiv.org/abs/2607.10995
  • 主分类:multimodal
  • 形态:method
  • 被引:0
  • 被引来源:Semantic Scholar
  • S2被引:0
  • 影响力被引:0
  • TLDR:This model matches optimization-based methods while delivering nearly 800x speedup, and surpasses the zero-shot performance of state-of-the-art generalizable models with markedly fewer parameters and reduced training/inference overhead, achieving an overall efficiency improvement.
  • 待LLM分类:否
  • 标题中文:AsySplat:用于长序列场景建模的高效非对称 3D Gaussian Splatting
  • TLDR中文:该模型在效果上匹配基于优化的方法,同时实现近 800 倍加速,并以明显更少的参数量和更低的训练/推理开销超越 SOTA 可泛化模型的零样本性能,整体效率显著提升。
  • 成熟度:research
  • 场景:3D重建、新视角合成、高斯泼溅
  • 来源文件
  • /inbox/tom/_candidates/2026-07-20-agent-memory-tool-use-candidates.json
  • [S2 enrich]