mlsanityData to 14 Oct 2026

A comparison of visual representations for real-world reinforcement learning in the context of vacuum gripping

arXiv 2503.02405 ↗First posted 4 Mar 2025cs.ROcited by 0

On the map: Deep reinforcement learning methods → RL for robot locomotion and manipulation

Abstract (arXiv)

When manipulating objects in the real world, we need reactive feedback policies that take into account sensor information to inform decisions. This study aims to determine how different encoders can be used in a reinforcement learning (RL) framework to interpret the spatial environment in the local surroundings of a robot arm. Our investigation focuses on comparing real-world vision with 3D scene inputs, exploring new architectures in the process. We built on the SERL framework, providing us with a sample efficient and stable RL foundation we could build upon, while keeping training times minimal. The results of this study indicate that spatial information helps to significantly outperform the visual counterpart, tested on a box picking task with a vacuum gripper. The code and videos of the evaluations are available at https://github.com/nisutte/voxel-serl.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in RL for robot locomotion and manipulation, one of 15 directions in Deep reinforcement learning methods. Builds on 32 of its 50 references that are in the map. First posted on arXiv 4 Mar 2025.

Builds on

32 of 50 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024SERL: A Software Suite for Sample-Efficient Robotic Reinforcement Learning29 Jan

Other references · 31

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  3. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,248
  4. 2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,885
  5. 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,568
  6. 2020Point Transformercited by 1,470
  7. 2015End-to-End Training of Deep Visuomotor Policiescited by 875
  8. 2018QT-Opt: Scalable Deep Reinforcement Learning for Vision-Based Robotic Manipulationcited by 743
  9. 2018Open3D: A Modern Library for 3D Data Processingcited by 721
  10. 2022R3M: A Universal Visual Representation for Robot Manipulationcited by 694
  11. 2020Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixelscited by 585
  12. 2020Reinforcement Learning with Augmented Datacited by 468
19 more
  1. 2023Efficient Online Reinforcement Learning with Offline Datacited by 363
  2. 2022Masked Visual Pre-training for Motor Controlcited by 248
  3. 2021Contact-GraspNet: Efficient 6-DoF Grasp Generation in Cluttered Scenescited by 218
  4. 2021A System for General In-Hand Object Re-Orientationcited by 165
  5. 2014Learning Rich Features from RGB-D Images for Object Detection and Segmentationcited by 127
  6. 2020The Ingredients of Real-World Robotic Reinforcement Learningcited by 117
  7. 2017Grasp Pose Detection in Point Cloudscited by 111
  8. 2022A Comprehensive Survey of Image Augmentation Techniques for Deep Learningcited by 97
  9. 2021RRL: Resnet as representation for Reinforcement Learningcited by 91
  10. 2024From Imitation to Refinement -- Residual RL for Precise Assemblycited by 90
  11. 2017Intel RealSense Stereoscopic Depth Camerascited by 83
  12. 2018Depth-aware CNN for RGB-D Segmentationcited by 74
  13. 2022Learning Generalizable Dexterous Manipulation from Human Grasp Affordancecited by 51
  14. 2021Generalization in Dexterous Manipulation via Geometry-Aware Multi-Task Learningcited by 50
  15. 2024Learning with 3D rotations, a hitchhiker's guide to SO(3)cited by 42
  16. 2022Frame Mining: a Free Lunch for Learning Robotic Manipulation from 3D Point Cloudscited by 32
  17. 2021InsertionNet -- A Scalable Solution for Insertioncited by 16
  18. 2022End-to-End Learning to Grasp via Sampling from Object Point Cloudscited by 15
  19. 2023On the Efficacy of 3D Point Cloud Reinforcement Learningcited by 11

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Benchmarking Action Spaces in Reinforcement Learning for Vision-based Robotic Manipulation
  2. 2024GAP-RL: Grasps As Points for RL Towards Dynamic Object Grasping ↗
  3. 2025SpatialVLA: Exploring Spatial Representations for Visual-Language-Action Model
  4. 2023Tracking Object Positions in Reinforcement Learning: A Metric for Keypoint Detection (extended version)
  5. 2023MoDem-V2: Visuo-Motor World Models for Real-World Robot Manipulation
  6. 2026Beyond Pixels: Learning Invariant Rewards for Real-World Robotics From a Few Demonstrations
  7. 2025Scaffolding Dexterous Manipulation with Vision-Language Models
  8. 2025ManiVID-3D: Generalizable View-Invariant Reinforcement Learning for Robotic Manipulation via Disentangled 3D Representations