Residual-NeRF: Learning Residual NeRFs for Transparent Object Manipulation

ICRA 2024

1Carnegie Mellon University,

Abstract

Transparent objects are ubiquitous in industry, pharmaceuticals, and households. Grasping and manipulating these objects is a significant challenge for robots. Existing methods have difficulty reconstructing complete depth maps for challenging transparent objects, leaving holes in the depth reconstruction. Recent work has shown neural radiance fields (NeRFs) work well for depth perception in scenes with transparent objects, and these depth maps can be used to grasp transparent objects with high accuracy. NeRF-based depth reconstruction can still struggle with especially challenging transparent objects and lighting conditions.

In this work, we propose Residual-NeRF, a method to improve depth perception and training speed for transparent objects. Robots often operate in the same area, such as a kitchen. By first learning a background NeRF of the scene without transparent objects to be manipulated, we reduce the ambiguity faced by learning the changes with the new object. We propose training two additional networks: a residual NeRF learns to infer residual RGB values and densities, and a Mixnet learns how to combine background and residual NeRFs. We contribute synthetic and real experiments that suggest Residual-NeRF improves depth perception of transparent objects. The results on synthetic data suggest Residual-NeRF outperforms the baselines with a 46.1% lower RMSE and a 29.5% lower MAE. Real-world qualitative experiments suggest Residual-NeRF leads to more robust depth maps with less noise and fewer holes.

Video

Visual Comparisons

Residual-NeRF
NeRF [Mildenhall 2020]
Residual-NeRF
Dex-NeRF [Ichnowski 2021]
Residual-NeRF
NeRF [Mildenhall 2020]
Residual-NeRF
Dex-NeRF [Ichnowski 2021]

Quantitative Evaluation

        Bowl   Drink Flat   Drink Up   Wine Flat  
Method RMSE ↓
NeRF   0.1900   0.6398   0.4361   0.2166  
Dex-NeRF 0.0365 0.1065 0.0699 0.0425
Residual-NeRF 0.0213 0.0234 0.0316 0.0388
        Bowl   Drink Flat   Drink Up   Wine Flat  
Method MAE ↓
NeRF   0.1453   0.3424   0.3062   0.1483  
Dex-NeRF 0.0195 0.0189 0.0255 0.0255
Residual-NeRF 0.0140 0.0138 0.0170 0.0142