mlsanityData to 14 Oct 2026

Exploiting Radiance Fields for Grasp Generation on Novel Synthetic Views

arXiv 2505.11467 ↗First posted 16 May 2025cs.RO cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Vision based robot manipulation uses cameras to capture one or more images of a scene containing the objects to be manipulated. Taking multiple images can help if any object is occluded from one viewpoint but more visible from another viewpoint. However, the camera has to be moved to a sequence of suitable positions for capturing multiple images, which requires time and may not always be possible, due to reachability constraints. So while additional images can produce more accurate grasp poses due to the extra information available, the time-cost goes up with the number of additional views sampled. Scene representations like Gaussian Splatting are capable of rendering accurate photorealistic virtual images from user-specified novel viewpoints. In this work, we show initial results which indicate that novel view synthesis can provide additional context in generating grasp poses. Our experiments on the Graspnet-1billion dataset show that novel views contributed force-closure grasps in addition to the force-closure grasps obtained from sparsely sampled real views while also improving grasp coverage. In the future we hope this work can be extended to improve grasp extraction from radiance fields constructed with a single input image, using for example diffusion models or generalizable radiance fields.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 31 of its 37 references that are in the map. First posted on arXiv 16 May 2025.

Builds on

31 of 37 references.

References · 31

  1. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
  2. 20233D Gaussian Splatting for Real-Time Radiance Field Renderingcited by 6,226
  3. 2020pixelNeRF: Neural Radiance Fields from One or Few Imagescited by 1,584
  4. 2022Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulationcited by 551
  5. 2023SplaTAM: Splat, Track & Map 3D Gaussians for Dense RGB-D SLAMcited by 388
  6. 2017Dex-Net 2.0: Deep Learning to Plan Robust Grasps with Synthetic Point Clouds and Analytic Grasp Metricscited by 264
  7. 20196-DOF GraspNet: Variational Grasp Generation for Object Manipulationcited by 242
  8. 2024A Survey on 3D Gaussian Splattingcited by 224
  9. 2021Contact-GraspNet: Efficient 6-DoF Grasp Generation in Cluttered Scenescited by 218
  10. 2021Coarse-to-Fine Q-attention: Efficient Learning for Visual Robotic Manipulation via Discretisationcited by 122
  11. 2021Dex-NeRF: Using a Neural Radiance Field to Grasp Transparent Objectscited by 119
  12. 2017Grasp Pose Detection in Point Cloudscited by 111
19 more
  1. 2021Synergies Between Affordance and Geometry: 6-DoF Grasp Detection via Implicit Representationscited by 108
  2. 2024ManiGaussian: Dynamic Gaussian Splatting for Multi-task Robotic Manipulationcited by 108
  3. 2019Grasping in the Wild:Learning 6DoF Closed-Loop Grasping from Low-Cost Demonstrationscited by 107
  4. 2022Vision Transformer for NeRF-Based View Synthesis from a Single Input Imagecited by 104
  5. 2021Volumetric Grasping Network: Real-time 6 DOF Grasp Detection in Cluttercited by 95
  6. 2018PointNetGPD: Detecting Grasp Configurations from Point Setscited by 80
  7. 2024GaussianGrasper: 3D Language Gaussian Splatting for Open-vocabulary Robotic Graspingcited by 78
  8. 2024How NeRFs and 3D Gaussian Splatting are Reshaping SLAM: a Surveycited by 72
  9. 2022GraspNeRF: Multiview-based 6-DoF Grasp Detection for Transparent and Specular Objects Using Generalizable NeRFcited by 67
  10. 2022Uncertainty Guided Policy for Active Robotic 3D Reconstruction using Neural Radiance Fieldscited by 54
  11. 2020MoreFusion: Multi-object Reasoning for 6D Pose Estimation from Volumetric Fusioncited by 46
  12. 2024GenNBV: Generalizable Next-Best-View Policy for Active 3D Reconstructioncited by 39
  13. 2023NViST: In the Wild New View Synthesis from a Single Image with Transformerscited by 19
  14. 2020Amodal 3D Reconstruction for Robotic Manipulation via Stability and Connectivitycited by 15
  15. 2023RGBGrasp: Image-based Object Grasping by Capturing Multiple Views during Robot Arm Movement with Neural Radiance Fieldscited by 13
  16. 2019Beyond Top-Grasps Through Scene Completioncited by 8
  17. 20246-DoF Grasp Pose Evaluation and Optimization via Transfer Learning from NeRFscited by 3
  18. 2020Probabilistic 3D Multilabel Real-time Mapping for Multi-object Manipulationcited by 2
  19. 2023Gradient based Grasp Pose Optimization on a NeRF that Approximates Grasp Successcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024GraspSplats: Efficient Manipulation with 3D Feature Splatting
  2. 2023GraspLDM: Generative 6-DoF Grasp Synthesis using Latent Diffusion Models
  3. 2023Language Embedded Radiance Fields for Zero-Shot Task-Oriented Grasping
  4. 2024Grasp Diffusion Network: Learning Grasp Generators from Partial Point Clouds with Diffusion Models in SO(3)xR3
  5. 2025VLAD-Grasp: Zero-shot Grasp Detection via Vision-Language Models
  6. 2024Constrained 6-DoF Grasp Generation on Complex Shapes for Improved Dual-Arm Manipulation
  7. 2025LensDFF: Language-enhanced Sparse Feature Distillation for Efficient Few-Shot Dexterous Manipulation
  8. 2025Grasp2Grasp: Vision-Based Dexterous Grasp Translation via Schr"odinger Bridges