NeRFにおけるサンプリング不要な障害物勾配と反応的計画
Sampling-free obstacle gradients and reactive planning in Neural Radiance Fields (NeRF)
事前学習済みNeRFに半径内の占有推定能力を追加してESDF近似を学習し、その微分から得た障害物勾配をRMPベースの回避方策に統合することで、サンプリング不要の高速な障害物回避計画を実現した。
著者: Michael Pantic, Cesar Cadena, Roland Siegwart, Lionel Ott
分類: cs.RO, cs.CV
原文アブストラクト
This work investigates the use of Neural implicit representations, specifically Neural Radiance Fields (NeRF), for geometrical queries and motion planning. We show that by adding the capacity to infer occupancy in a radius to a pre-trained NeRF, we are effectively learning an approximation to a Euclidean Signed Distance Field (ESDF). Using backward differentiation of the augmented network, we obtain an obstacle gradient that is integrated into an obstacle avoidance policy based on the Riemannian Motion Policies (RMP) framework. Thus, our findings allow for very fast sampling-free obstacle avoidance planning in the implicit representation.
