⚡ 本页包含 AI 生成的分析内容，仅供参考
本文提出一种基于稀疏混合专家（MoE）的NeRF-SLAM处理器，用于移动空间计算。该处理器实现了303.5mW低功耗下的实时稠密3D建图，解决了传统SLAM处理器只支持稀疏特征点且需要额外后处理的问题。
Recently, spatial computing has become popular in mobile devices, such as autonomous robots and augmented reality (AR) glasses [1], and it enables cyber-physical interaction through accurate user position and 3D geometric information of the surrounding environment obtained with the simultaneous localization and mapping (SLAM) algorithm. Previous SLAM processors [2-5] accelerated mapping and tracking, but they supported few (<5K) point features, and required additional post processing (volumetric fusion [6]) for dense 3D map acquisition. Real-time SLAM processing is impossible on memoryconstrained mobile devices due to the large (>60MB) dense 3D map representation which stores color/distance values in high resolution (<4 cm) voxel. A neural radiance fields
Gwangtae Park1, Seokchan Song1, Haoyang Sang1, Dongseok Im1,
Donghyeon Han2, Sangyeob Kim1, Hongseok Lee1, Hoi-Jun Yoo1 Korea Advanced Institute of Science and Technology, Daejeon, Korea 2 Massachusetts Institute of Technology, Cambridge, MA 1