Exemplar Fine-Tuning for 3D Human Model Fitting Towards In-the-Wild 3D Human Pose Estimation
Open accessThe DOI matches its Crossref record; 10 of the 10 references it cites were found in a registry or an index.
Abstract
Differently from 2D image datasets such as COCO, largescale human datasets with 3D ground-truth annotations are very difficult to obtain in the wild. In this paper, we address this problem by augmenting existing 2D datasets with high-quality 3D pose fits. Remarkably, the resulting annotations are sufficient to train from scratch 3D pose regressor networks that outperform the current state-of-the-art on in the-wild benchmarks such as 3DPW. Additionally, training on our augmented data is straightforward as it does not require to mix multiple and incompatible 2D and 3D datasets or to use complicated network architectures and training procedures. This simplified pipeline affords additional improvements, including injecting extreme crop augmentations to better reconstruct highly truncated people, and incorporating auxiliary inputs to improve 3D pose estimation accuracy. It also reduces the dependency on 3D datasets such as H36M that have restrictive licenses. We also use our method to introduce new benchmarks for the study of real-world challenges such as occlusions, truncations, and rare body poses. In order to obtain such high quality 3D pseudo-annotations, inspired by progress in internal learning, we introduce Exemplar Fine-Tuning (EFT). EFT combines the re-projection accuracy of fitting methods like SMPLify with a 3D pose prior implicitly captured by a pre-trained 3D pose regressor network. We show that EFT produces 3D annotations that result in better downstream performance and are qualitatively preferable in an extensive human-based assessment.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946
Realtime Multi-person 2D Pose Estimation Using Part Affinity Fields
Cao, Z., Simon, T., Wei, S.-E., & Sheikh, Y. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.143 · Cited by 7,539 · cited 1×
“Deep learning has significantly advanced 2D pose recognition [10, 9, 63, 44, 66, 57], facilitating the more challenging task of 3D reconstruction [59, 60, 40, 49, 48, 45, 50, 27, 35, 65, 11, 58, 56, 13, 43, 12], which is our focus.”
BackgroundDeep High-Resolution Representation Learning for Human Pose Estimation
Sun, K., Xiao, B., Liu, D., & Wang, J. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00584 · Cited by 5,984
Stacked Hourglass Networks for Human Pose Estimation
Newell, A., Yang, K., & Deng, J. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46484-8_29 · Cited by 5,634 · cited 1×
“Deep learning has significantly advanced 2D pose recognition [10, 9, 62, 43, 65, 56], facilitating the more challenging task of 3D reconstruction [58, 59, 39, 48, 47, 44, 49, 26, 34, 64, 11, 57, 55, 13, 42, 12], which is our focus.”
BackgroundOpenPose: Realtime Multi-Person 2D Pose Estimation Using Part Affinity Fields
Cao, Z., Hidalgo, G., Simon, T., Wei, S.-E., & Sheikh, Y. (2021) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2019.2929257 · Cited by 5,536
Loper, M., Mahmood, N., Romero, J., Pons-Moll, G., & J, B. M. (2015) · ACM Transactions on Graphics · doi:10.1145/2816795.2818013 · Cited by 3,734 · cited 3×
“Following [6, 37, 25, 46, 45], the human is represented by a parametric model that spans variation in both body shapes and motions.”
MethodHuman3.6M: Large Scale Datasets and Predictive Methods for 3D Human Sensing in Natural Environments
Ionescu, C., Papava, D., Olaru, V., & Sminchisescu, C. (2014) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2013.248 · Cited by 3,653 · cited 5×
“state-of-the-art approaches [26, 27, 33, 32] rely on SMPL fits on the H36M dataset [21, 20] which, due to licensing restrictions, make reproducibility problematic.”
BackgroundMethodInfluentialExpressive Body Capture: 3D Hands, Face, and Body From a Single Image
Pavlakos, G., Choutas, V., Ghorbani, N., Bolkart, T., Osman, A. A., Tzionas, D., & Black, M. J. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.01123 · Cited by 1,781 · cited 19×
“For the network Φ w ∗ , we use HMR [27] pre-trained using SPIN [34].”
MethodInfluentialSTAR: Sparse Trained Articulated Human Body Regressor
Osman, A. A. A., Bolkart, T., & Black, M. J. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58539-6_36 · Cited by 252 · cited 3×
“Fitting-based methods assume a 3D body model such as SCAPE [6], SMPL [38], SMPL-X[47], and STAR [46], and use an optimization algorithm to fit it to the 2D observations.”
MethodThree-D Safari: Learning to Estimate Zebra Pose, Shape, and Texture From Images “In the Wild”
Zuffi, S., Kanazawa, A., Berger-Wolf, T., & Black, M. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00546 · Cited by 139 · cited 1×
“While similar techniques have been used in [60, 3, 73] to improve test-time performance, they have not been rigorously assessed against classical optimization approaches such as SMPLify, nor they have been used to generate pseudo-labels.”
Method
Cited by
Moon, G., & Lee, K. M. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58571-6_44 · Cited by 377
Humans in 4D: Reconstructing and Tracking Humans with Transformers
Goel, S., Pavlakos, G., Rajasegaran, J., Kanazawa, A., & Malik, J. (2023) · 2023 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.01358 · Cited by 272
CLIFF: Carrying Location Information in Full Frames into Human Pose and Shape Estimation
Li, Z., Liu, J., Zhang, Z., Xu, S., & Yan, Y. (2022) · Lecture Notes in Computer Science · doi:10.1007/978-3-031-20065-6_34 · Cited by 268
A Systematic Review of Recent Deep Learning Approaches for 3D Human Pose Estimation
El Kaid, A., & Baïna, K. (2023) · Journal of Imaging · doi:10.3390/jimaging9120275 · Cited by 44
Metadata
- Authors
- Joo, H., Neverova, N., & Vedaldi, A.
- Venue
- 2021 International Conference on 3D Vision (3DV)
- Publisher
- IEEE
- Published
- 2021-12
- Pages
- 42-52
- OpenAlex
- W4205116876
- Cited by
- 187
- References
- 10
- FWCI
- 12.04
- Type
- conference paper
- Language
- en
- Open access
- green
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Exemplar Fine-Tuning for 3D Human Model Fitting Towards In-the-Wild 3D Human Pose Estimation. https://app.cochat.ai/papers/10.1109/3dv53792.2021.00015
- The paper
- Joo, H., Neverova, N., & Vedaldi, A. (2021). Exemplar Fine-Tuning for 3D Human Model Fitting Towards In-the-Wild 3D Human Pose Estimation. 2021 International Conference on 3D Vision (3DV). https://doi.org/10.1109/3dv53792.2021.00015
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.