mlsanityData to 14 Oct 2026

Supervised Training of Dense Object Nets using Optimal Descriptors for Industrial Robotic Applications

arXiv 2102.08096 ↗First posted 16 Feb 2021cs.RO cs.CVcited by 4

Not on the map. Why?

Abstract (arXiv)

Dense Object Nets (DONs) by Florence, Manuelli and Tedrake (2018) introduced dense object descriptors as a novel visual object representation for the robotics community. It is suitable for many applications including object grasping, policy learning, etc. DONs map an RGB image depicting an object into a descriptor space image, which implicitly encodes key features of an object invariant to the relative camera pose. Impressively, the self-supervised training of DONs can be applied to arbitrary objects and can be evaluated and deployed within hours. However, the training approach relies on accurate depth images and faces challenges with small, reflective objects, typical for industrial settings, when using consumer grade depth cameras. In this paper we show that given a 3D model of an object, we can generate its descriptor space image, which allows for supervised training of DONs. We rely on Laplacian Eigenmaps (LE) to embed the 3D model of an object into an optimally generated space. While our approach uses more domain knowledge, it can be efficiently applied even for smaller and reflective objects, as it does not rely on depth information. We compare the training methods on generating 6D grasps for industrial objects and show that our novel supervised training approach improves the pick-and-place performance in industry-relevant tasks.

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Builds on 11 of its 26 references that are in the map. First posted on arXiv 16 Feb 2021.

Builds on

11 of 26 references.

References · 11

  1. 2014Fully Convolutional Networks for Semantic Segmentationcited by 4,859
  2. 2018DeepIM: Deep Iterative Matching for 6D Pose Estimationcited by 283
  3. 2019DPOD: 6D Pose Object Detector and Refinercited by 189
  4. 2018Dense Object Nets: Learning Dense Visual Object Descriptors By and For Robotic Manipulationcited by 142
  5. 2019Self-Supervised Correspondence in Visuomotor Policy Learningcited by 107
  6. 2016Universal Correspondence Networkcited by 82
  7. 2020Keypoints into the Future: Self-Supervised Correspondence in Model-Based Reinforcement Learningcited by 70
  8. 2020Learning Rope Manipulation Policies Using Dense Object Descriptors Trained on Synthetic Depth Datacited by 51
  9. 2018The Vector Heat Methodcited by 31
  10. 2020S3K: Self-Supervised Semantic Keypoints for Robotic Manipulation via Multi-View Consistencycited by 28
  11. 2019Refining 6D Object Pose Predictions using Abstract Render-and-Comparecited by 8

Built on by

Cited by year

Citing papers per year: 2022: 1; 2023: 1; 2024: 2.

4 papers.

  1. 2022Efficient and Robust Training of Dense Object Nets for Multi-Object Robot Manipulationcited by 2
  2. 2023Learning Sim-to-Real Dense Object Descriptors for Robotic Manipulationcited by 2
  3. 2024Cycle-Correspondence Loss: Learning Dense View-Invariant Visual Features from Unlabeled and Unordered RGB Imagescited by 1
  4. 2024Twin Trigger Generative Networks for Backdoor Attacks against Object Detectioncited by 0