Material Links
MEDIA ANALYTICS
PROJECTS
PEOPLE
PUBLICATIONS
PATENTS
Deep Supervision With Shape Concepts for Occlusion-Aware 3D Object Parsing
We propose a deep CNN architecture to localize object semantic parts in 2D images and 3D space while inferring their visibility states given a single RGB image. We exploit domain knowledge to regularize the network by deeply supervising its hidden layers. In doing so, we sequentially infer a causal sequence of intermediate concepts. We render 3D object CAD models to generate large-scale synthetic data and simulate challenging occlusion configurations between objects. The utility of our deep supervision is demonstrated by state-of-the-art performances on real image benchmarks for 2D and 3D keypoint localization and instance segmentation.
Collaborators: Chi Li, Zeeshan Zia, Quoc-Huy Tran, Gregory D. Hager, Manmohan Chandraker
Dataset
Rendered Images (Car)
This package contains the rendered images of cars that were used in the papers: “Deep Supervision with Shape Concepts for Occlusion-Aware 3D Object Parsing, Chi Li, M. Zeeshan Zia, Quoc-Huy Tran, Xiang Yu, Gregory D. Hager, Manmohan Chandraker, IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017” and “Deep Supervision with Intermediate Concepts, Chi Li, M. Zeeshan Zia, Quoc-Huy Tran, Xiang Yu, Gregory D. Hager, Manmohan Chandraker, IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2018”. To obtain the data, please download the below ZIP files and license agreement.
Important: The following instructions must be followed exactly:
- The license agreement must be signed by the appropriate manager responsible for enforcing copyright. For a student, it may be the faculty advisor. For a professor, it may be the department chair. For a researcher, it may be the lab or company manager.
- All three pages of the agreement must be scanned and emailed.
- Please use your official university or company email address.
You will receive a password to unlock the data in your official email address.
- ZIP file for non-occluded cars
- ZIP file for occluded cars
- License agreement
- Email address to send the signed agreement
Please cite the below papers if you use the above data in your work:
@inproceedings{li2017deep, title={Deep supervision with shape concepts for occlusion-aware 3d object parsing}, author={Li, Chi and Zeeshan Zia, M and Tran, Quoc-Huy and Yu, Xiang and Hager, Gregory D and Chandraker, Manmohan}, booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition}, pages={5465--5474}, year={2017} } @article{li2018deep, title={Deep supervision with intermediate concepts}, author={Li, Chi and Zia, M Zeeshan and Tran, Quoc-Huy and Yu, Xiang and Hager, Gregory D and Chandraker, Manmohan}, journal={IEEE transactions on pattern analysis and machine intelligence}, year={2018}, publisher={IEEE} }
Deep Supervision Publications
Nothing Found
Sorry, no posts matched your criteria