mlsanityData to 14 Oct 2026

OSSID: Online Self-Supervised Instance Detection by (and for) Pose Estimation

arXiv 2201.07309 ↗First posted 18 Jan 2022cs.CV cs.AI cs.ROcited by 8

Not on the map. Why?

Abstract (arXiv)

Real-time object pose estimation is necessary for many robot manipulation algorithms. However, state-of-the-art methods for object pose estimation are trained for a specific set of objects; these methods thus need to be retrained to estimate the pose of each new object, often requiring tens of GPU-days of training for optimal performance. In this paper, we propose the OSSID framework, leveraging a slow zero-shot pose estimator to self-supervise the training of a fast detection algorithm. This fast detector can then be used to filter the input to the pose estimator, drastically improving its inference speed. We show that this self-supervised training exceeds the performance of existing zero-shot detection methods on two widely used object pose estimation and detection datasets, without requiring any human annotations. Further, we show that the resulting method for pose estimation has a significantly faster inference speed, due to the ability to filter out large parts of the image. Thus, our method for self-supervised online learning of a detector (trained using pseudo-labels from a slow pose estimator) leads to accurate pose estimation at real-time speeds, without requiring human annotations. Supplementary materials and code can be found at https://georgegu1997.github.io/OSSID/

Key facts. Cited by 8 papers in the map (data through 14 Oct 2026). Builds on 33 of its 61 references that are in the map. First posted on arXiv 18 Jan 2022.

Builds on

33 of 61 references.

References · 33

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  3. 2017Mask R-CNNcited by 7,277
  4. 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,897
  5. 2019On the Convergence of Adam and Beyondcited by 841
  6. 2017PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenescited by 775
  7. 2018Domain Adaptive Faster R-CNN for Object Detection in the Wildcited by 443
  8. 2019DenseFusion: 6D Object Pose Estimation by Iterative Dense Fusioncited by 355
  9. 2020CosyPose: Consistent multi-view multi-object 6D pose estimationcited by 287
  10. 2020Frustratingly Simple Few-Shot Object Detectioncited by 282
  11. 2018Fast Online Object Tracking and Segmentation: A Unifying Approachcited by 272
  12. 2018Few-shot Object Detection via Feature Reweightingcited by 265
21 more
  1. 2018BOP: Benchmark for 6D Object Pose Estimationcited by 254
  2. 2019Few-Shot Object Detection with Attention-RPN and Multi-Relation Detectorcited by 206
  3. 2020Few-Shot Segmentation Without Meta-Learning: A Good Transductive Inference Is All You Need?cited by 135
  4. 2019Progressive Domain Adaptation for Object Detectioncited by 134
  5. 2019Multi-adversarial Faster-RCNN for Unrestricted Object Detectioncited by 112
  6. 2019A Robust Learning Approach to Domain Adaptive Object Detectioncited by 109
  7. 2020Cross-domain Detection via Graph-induced Prototype Alignmentcited by 109
  8. 2021MeGA-CDA: Memory Guided Attention for Category-Aware Unsupervised Domain Adaptive Object Detectioncited by 106
  9. 2021Unsupervised Domain Adaptation of Object Detectors: A Surveycited by 93
  10. 2019Self-Training and Adversarial Background Regularization for Unsupervised Domain Adaptive One-Stage Object Detectioncited by 83
  11. 2019BlenderProccited by 77
  12. 2019Photorealistic Image Synthesis for Object Instance Detectioncited by 66
  13. 2019Automatic adaptation of object detectors to new domains using self-trainingcited by 66
  14. 2019Multi-path Learning for Object Pose Estimation Across Domainscited by 52
  15. 2019Pose from Shape: Deep Pose Estimation for Arbitrary 3D Objectscited by 39
  16. 2017Going Further with Point Pair Featurescited by 28
  17. 2021ZePHyR: Zero-shot Pose Hypothesis Ratingcited by 26
  18. 2018Target Driven Instance Detectioncited by 14
  19. 2019Online Object Representations with Contrastive Learningcited by 13
  20. 2020A Hybrid Approach for 6DoF Pose Estimationcited by 13
  21. 2017A Self-supervised Learning System for Object Detection using Physics Simulation and Multi-view Pose Estimationcited by 11

Built on by

Cited by year

Citing papers per year: 2022: 3; 2023: 3; 2024: 2.

8 papers.

  1. 2022Gen6D: Generalizable Model-Free 6-DoF Object Pose Estimation from RGB Imagescited by 121
  2. 2022Deep Unsupervised Domain Adaptation: A Review of Recent Advances and Perspectivescited by 98
  3. 2024EgoLifter: Open-world 3D Segmentation for Egocentric Perceptioncited by 47
  4. 2023Challenges for Monocular 6D Object Pose Estimation in Roboticscited by 21
  5. 2023VoxDet: Voxel Learning for Novel Instance Detectioncited by 8
  6. 2023Open Challenges for Monocular Single-shot 6D Object Pose Estimationcited by 2
  7. 2024Diffusion-Driven Self-Supervised Learning for Shape Reconstruction and Pose Estimationcited by 1
  8. 2022One-Shot General Object Localizationcited by 0