mlsanityData to 7 Oct 2026

PR-RRN: Pairwise-Regularized Residual-Recursive Networks for Non-rigid Structure-from-Motion

arXiv 2108.07506 ↗First posted 17 Aug 2021cs.CVcited by 9

Not on the map. Why?

Abstract (arXiv)

We propose PR-RRN, a novel neural-network based method for Non-rigid Structure-from-Motion (NRSfM). PR-RRN consists of Residual-Recursive Networks (RRN) and two extra regularization losses. RRN is designed to effectively recover 3D shape and camera from 2D keypoints with novel residual-recursive structure. As NRSfM is a highly under-constrained problem, we propose two new pairwise regularization to further regularize the reconstruction. The Rigidity-based Pairwise Contrastive Loss regularizes the shape representation by encouraging higher similarity between the representations of high-rigidity pairs of frames than low-rigidity pairs. We propose minimum singular-value ratio to measure the pairwise rigidity. The Pairwise Consistency Loss enforces the reconstruction to be consistent when the estimated shapes and cameras are exchanged between pairs. Our approach achieves state-of-the-art performance on CMU MOCAP and PASCAL3D+ dataset.

Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). Builds on 15 of its 52 references that are in the map. First posted on arXiv 17 Aug 2021.

Builds on

15 of 52 references.

References · 15

  1. 2018Representation Learning with Contrastive Predictive Codingcited by 7,662
  2. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,482
  3. 2019Contrastive Multiview Codingcited by 1,297
  4. 2018On the Continuity of Rotation Representations in Neural Networkscited by 1,186
  5. 2019Unsupervised 3D Pose Estimation with Geometric Self-Supervisioncited by 95
  6. 2020An Analysis of SVD for Deep Rotation Estimationcited by 85
  7. 2020Info3D: Representation Learning on 3D Objects using Mutual Information Maximization and Contrastive Learningcited by 68
  8. 2019C3DPO: Canonical 3D Pose Networks for Non-Rigid Structure From Motioncited by 60
  9. 2017Pixel Recursive Super Resolutioncited by 44
  10. 2018Scalable Dense Non-rigid Structure-from-Motion: A Grassmannian Perspectivecited by 19
  11. 2018Structure from Recurrent Motion: From Rigidity to Recurrencycited by 16
  12. 2021Self-Supervised Visibility Learning for Novel View Synthesiscited by 14
3 more
  1. 2020Procrustean Regression Networks: Learning 3D Structure of Non-Rigid Objects from 2D Annotationscited by 13
  2. 2019Jumping Manifolds: Geometry Aware Dense Non-Rigid Structure from Motioncited by 12
  3. 2017Spatial-Temporal Union of Subspaces for Multi-body Non-rigid Structure-from-Motioncited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 1; 2022: 4; 2023: 2; 2024: 1; 2025: 0; 2026: 1.

9 papers.

  1. 2022State of the Art in Dense Monocular Non-Rigid 3D Reconstructioncited by 30
  2. 2024Recent Trends in 3D Reconstruction of General Non-Rigid Scenescited by 28
  3. 2023CAMM: Building Category-Agnostic and Animatable 3D Models from Monocular Videoscited by 10
  4. 2022Deep Non-rigid Structure-from-Motion: A Sequence-to-Sequence Translation Perspectivecited by 8
  5. 2022MHR-Net: Multiple-Hypothesis Reconstruction of Non-Rigid Shapes from 2D Viewscited by 7
  6. 2023Unsupervised 3D Pose Estimation with Non-Rigid Structure-from-Motion Modelingcited by 7
  7. 2021End-to-End Learning of Multi-category 3D Pose and Shape Estimationcited by 1
  8. 2022Piecewise Planar Hulls for Semi-Supervised Learning of 3D Shape and Pose from 2D Imagescited by 0
  9. 2026SSR: A Training-Free Approach for Streaming 3D Reconstructioncited by 0