Car make and model recognition using 3D curve alignment

Hsiao, Edward; Sinha, Sudipta N.; Ramnath, Krishnan; Baker, Simon; Zitnick, Larry; Szeliski, Richard

doi:10.1109/wacv.2014.6836126

Cited by 29 publications

(15 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the past, textureless objects have often been handled by template-matching [11,12]. Image edges then become the dominant information source [21,26], and researchers have developed strategies based on different distances, such as the Hausdorff [15] and the Chamfer [25,13] ones, to match the 3D model against the input image. While effective for poorly-textured objects, these techniques often fail in the presence of mild occlusions and cluttered background.…”

Section: Related Workmentioning

confidence: 99%

Segmentation-Driven 6D Object Pose Estimation

Hugonot

Fua

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

316

242

View full text Add to dashboard Cite

The most recent trend in estimating the 6D pose of rigid objects has been to train deep networks to either directly regress the pose from the image or to predict the 2D locations of 3D keypoints, from which the pose can be obtained using a PnP algorithm. In both cases, the object is treated as a global entity, and a single pose estimate is computed. As a consequence, the resulting techniques can be vulnerable to large occlusions.In this paper, we introduce a segmentation-driven 6D pose estimation framework where each visible part of the objects contributes a local pose prediction in the form of 2D keypoint locations. We then use a predicted measure of confidence to combine these pose candidates into a robust set of 3D-to-2D correspondences, from which a reliable pose estimate can be obtained. We outperform the state-ofthe-art on the challenging Occluded-LINEMOD and YCB-Video datasets, which is evidence that our approach deals well with multiple poorly-textured objects occluding each other. Furthermore, it relies on a simple enough architecture to achieve real-time performance. Input CNNResult

show abstract

Section: Related Workmentioning

confidence: 99%

Segmentation-Driven 6D Object Pose Estimation

Hugonot

Fua

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

316

242

View full text Add to dashboard Cite

show abstract

“…(Some preliminary results on incorporating taillights can be found in [10].) We will explore extensions that combine appearance and geometric features with the goal of improving accuracy in detection, pose-estimation and make and model recognition of a car from a single image.…”

Section: Resultsmentioning

confidence: 99%

“…The eight models span various car categories such as sedan, SUV, hatchback, mini-van and truck, and are shown in Figure 9 (Appendix A) of our supplementary [10]. We avoided images taken against uniform backgrounds, since we wanted to test our system operating "in the wild", with cars shot in front of varied textured backgrounds from different viewpoints and with lots of body reflections.…”

Section: Discussionmentioning

confidence: 99%

“…Finally, the focal length can be determined from the rotation matrix and the original vanishing points. We provide more detailed mathematical derivation in our supplementary material [10].…”

Section: Ellipse-based Pose Estimationmentioning

confidence: 99%

“…Since we have built a 3D model for each of the cars, we use four known 3D points corresponding to the wheel rim tops and bottoms (for each car model) to compute a full 3D pose estimate. The details here are even simpler than the model-free pose estimation (essentially using a direct linear transform) and are also described in the supplementary materials [10].…”

Section: Ellipse-based Pose Estimationmentioning

confidence: 99%

See 2 more Smart Citations

Car make and model recognition using 3D curve alignment

Ramnath¹,

Sinha²,

Szeliski³

et al. 2014

IEEE Winter Conference on Applications of Computer Vision

Self Cite

View full text Add to dashboard Cite

We present a new approach for recognizing the make and model of a car from a single image. While most previous methods are restricted to fixed or limited viewpoints, our system is able to verify a car's make and model from an arbitrary view. Our model consists of 3D space curves obtained by backprojecting image curves onto silhouettebased visual hulls and then refining them using three-view curve matching. These 3D curves are then matched to 2D image curves using a 3D view-based alignment technique. We present two different methods for estimating the pose of a car, which we then use to initialize the 3D curve matching. Our approach is able to verify the exact make and model of a car over a wide range of viewpoints in cluttered scenes.

show abstract

View-Based 3D Objects Recognition with Expectation Propagation Learning

Bertrand

Al-Osaimi

Bouguila

2016

Advances in Visual Computing

View full text Add to dashboard Cite

View-based 3D Objects Recognition with Expectation Propagation LearningIn this thesis, we present an improvement on the Expectation Propagation learning framework, specifically various enhancements on both speed and accuracy. We use this enhanced EP learning with the Inverted Dirichlet mixture model as well as the Dirichlet mixture model, to implement an algorithm to recognize 3D objects. Those objects are in our case from a view-based 3D models database that we have assembled. Following specific rules determined by analyzing the results of our tests, we've been able to get good recognition rates. Experimental results are presented with different object classes by comparing recognition rates and confidence level, according to different tuning parameters we're able to refine towards specific classes for better specialized accuracy. iv ACKNOWLEDGMENTSFirst, I would like to thank my supervisor Dr. Nizar Bouguila, for the many hours spent with me at Concordia and on Skype, answering my questions and guiding me all along the way.Moreover, I want to express my gratitude to Dr. Taoufik Bdiri for giving me some pointers and a nice trick regarding an initialization algorithm.I would also like to thank my friends and family, of course, for their support and continuous encouragement, and especially my father, Paul Bertrand, for his precious math help in particular. v

show abstract

Car make and model recognition using 3D curve alignment

Cited by 29 publications

References 29 publications

Segmentation-Driven 6D Object Pose Estimation

Segmentation-Driven 6D Object Pose Estimation

Car make and model recognition using 3D curve alignment

View-Based 3D Objects Recognition with Expectation Propagation Learning

Contact Info

Product

Resources

About