Filters








23,910 Hits in 3.4 sec

Viewpoint Adaptation for Rigid Object Detection [article]

Patrick Wang and Kenneth Morton and Peter Torrione and Leslie Collins
2017 arXiv   pre-print
This work has the potential to improve detection performance for cameras viewing objects from arbitrary viewpoints, while simplifying data collection and feature extraction.  ...  In this paper, we present a viewpoint adaptation algorithm that allows a trained single-view object detector to be adapted to a new, distinct viewpoint.  ...  More generally, it is desirable to be able to adapt any rigid object detector trained for one viewpoint to another distinct viewpoint.  ... 
arXiv:1702.07451v1 fatcat:4rbja3i2qvcvbdq6km56ria2am

Horizontal fixation point oscillation and simulated viewpoint oscillation both increase vection in depth

S. Palmisano, J. Kim, T. C. A. Freeman
2012 Journal of Vision  
Neither vection advantage could be explained in terms of differences in perceived scene rigidity or motion adaptation.  ...  Our findings also provided little support for the notion that pursuit eye-movements were essential for the simulated viewpoint oscillation advantage for vection (since observers successfully fixated a  ...  Little support was found for the increased perceived rigidity explanation of the simulated viewpoint oscillation advantage for vection.  ... 
doi:10.1167/12.12.15 pmid:23184234 fatcat:st5hflcpdncwfiieinekbyevrq

Horizontal fixation point oscillation and simulated viewpoint oscillation both increase vection in depth

S. Palmisano, J. Kim, T. Freeman
2012 Journal of Vision  
Neither vection advantage could be explained in terms of differences in perceived scene rigidity or motion adaptation.  ...  Our findings also provided little support for the notion that pursuit eye-movements were essential for the simulated viewpoint oscillation advantage for vection (since observers successfully fixated a  ...  Little support was found for the increased perceived rigidity explanation of the simulated viewpoint oscillation advantage for vection.  ... 
doi:10.1167/12.9.196 fatcat:34e6szjdxjf2tjo6k7up2pgxwa

Few-Shot Viewpoint Estimation [article]

Hung-Yu Tseng, Shalini De Mello, Jonathan Tremblay, Sifei Liu, Stan Birchfield, Ming-Hsuan Yang, Jan Kautz
2019 arXiv   pre-print
Viewpoint estimation for known categories of objects has been improved significantly thanks to deep networks and large datasets, but generalization to unknown categories is still very challenging.  ...  We design a novel framework to successfully train viewpoint networks for new categories with few examples (10 or less).  ...  Appendix B Additional Experiments Non-rigid objects. As a preliminary investigation, we show the potential of the proposed method on detecting semantic keypoints of non-rigid objects.  ... 
arXiv:1905.04957v2 fatcat:q66sirl6yrfm5gqm7ti7utvmqu

Adaptive region pooling for object detection

Yi-Hsuan Tsai, Onur C. Hamsici, Ming-Hsuan Yang
2015 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
Learning models for object detection is a challenging problem due to the large intra-class variability of objects in appearance, viewpoints, and rigidity.  ...  We also show that object information can be transferred through exemplars for detected objects.  ...  Non-rigid object models [19] and 3D CAD models [3, 28] can also be used to generalize the application of transferring rigid and non-rigid object information.  ... 
doi:10.1109/cvpr.2015.7298673 dblp:conf/cvpr/TsaiHY15 fatcat:p65xbf3mxfaxdnjgyqsl5dqjte

Retrieving Objects From Videos Based On Affine Regions

Luc Van Gool, Tinne Tuytelaars, Vittorio Ferrari
2004 Zenodo  
Third, non-rigid deformations are taken into account. The method can extend any viewpoint invariant region extractor. In [9] another region-based system for video object retrieval is presented.  ...  This is a simplified version of our method [3] adapted for the application. It was first used for object recognition in still images.  ... 
doi:10.5281/zenodo.38621 fatcat:6b646twiojcn3gxbxenov7ir5q

3D Affine: An Embedding of Local Image Features for Viewpoint Invariance Using RGB-D Sensor Data

Hamdi Sahloul, Shouhei Shirafuji, Jun Ota
2019 Sensors  
Invariance to such viewpoint changes is essential for numerous applications, including wide baseline matching, 6D pose estimation, and object reconstruction.  ...  Using synthetic and real-world objects, we evaluated the viewpoint invariance of various detectors and descriptors, for both standalone and embedded approaches.  ...  adaptations.  ... 
doi:10.3390/s19020291 fatcat:4zcccqhq4jgvdniy6l7mexdzby

Viewpoints and keypoints

Shubham Tulsiani, Jitendra Malik
2015 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
We characterize the problem of pose estimation for rigid objects in terms of determining viewpoint to explain coarse pose and keypoint prediction to capture the finer details.  ...  We present Convolutional Neural Network based architectures for these and demonstrate that leveraging viewpoint estimates can substantially improve local appearance based keypoint predictions.  ...  We gratefully acknowledge NVIDIA corporation for the donation of Tesla GPUs for this research.  ... 
doi:10.1109/cvpr.2015.7298758 dblp:conf/cvpr/TulsianiM15 fatcat:3bhc6d7utnfefenzs3aokglzta

Cylindrical Convolutional Networks for Joint Object Detection and Viewpoint Estimation [article]

Sunghun Joung, Seungryong Kim, Hanjae Kim, Minsu Kim, Ig-Jae Kim, Junghyun Cho, Kwanghoon Sohn
2020 arXiv   pre-print
This does not account for the fact that objects in a 2D space are a projection of 3D ones, and thus they have limited ability to severe object viewpoint changes.  ...  Our experiments demonstrate the effectiveness of the cylindrical convolutional networks on joint object detection and viewpoint estimation.  ...  This research was supported by R&D program for Advanced Integrated-intelligence for Identification (AIID) through the National Research Foundation of KOREA (NRF) funded by Ministry of Science and ICT (  ... 
arXiv:2003.11303v1 fatcat:kb5gepiikjdsli73pt36fkfv7q

Lifting Object Detection Datasets into 3D [article]

Joao Carreira, Sara Vicente, Lourdes Agapito, Jorge Batista
2016 arXiv   pre-print
We show that our method is able to produce convincing per-object 3D reconstructions and to accurately estimate cameras viewpoints on one of the most challenging existing object-category detection datasets  ...  Our proposed algorithm first estimates camera viewpoint using rigid structure-from-motion and then reconstructs object shapes by optimizing over visual hull proposals guided by loose within-class shape  ...  We adapt it to operate on category detection imagery proposing a novel formulation that we denote imprinted visual hull reconstruction.  ... 
arXiv:1503.06465v2 fatcat:24uzuqrs4rbzfpmm3rknhghw4y

Category-Specific Object Reconstruction from a Single Image [article]

Abhishek Kar, Shubham Tulsiani, João Carreira, Jitendra Malik
2015 arXiv   pre-print
At the core of our approach are deformable 3D models that can be learned from 2D annotations available in existing object detection datasets, that can be driven by noisy automatic object segmentations  ...  and which we complement with a bottom-up module for recovering high-frequency shape details.  ...  We gratefully acknowledge NVIDIA corporation for the donation of Tesla GPUs for this research.  ... 
arXiv:1411.6069v2 fatcat:g4f5xqheprh67p5gqvn7rlfi4u

Lifting Object Detection Datasets into 3D

Joao Carreira, Sara Vicente, Lourdes Agapito, Jorge Batista
2016 IEEE Transactions on Pattern Analysis and Machine Intelligence  
We show that our method is able to produce convincing per-object 3D reconstructions and to accurately estimate cameras viewpoints on one of the most challenging existing object-category detection datasets  ...  Our proposed algorithm first estimates camera viewpoint using rigid structure-from-motion and then reconstructs object shapes by optimizing over visual hull proposals guided by loose within-class shape  ...  We adapt it to operate on category detection imagery proposing a novel formulation that we denote imprinted visual hull reconstruction.  ... 
doi:10.1109/tpami.2015.2435707 pmid:27295458 fatcat:sbrzyoo33zcdhfvewg2uur2vua

Viewpoints and Keypoints [article]

Shubham Tulsiani, Jitendra Malik
2015 arXiv   pre-print
We characterize the problem of pose estimation for rigid objects in terms of determining viewpoint to explain coarse pose and keypoint prediction to capture the finer details.  ...  We present Convolutional Neural Network based architectures for these and demonstrate that leveraging viewpoint estimates can substantially improve local appearance based keypoint predictions.  ...  We gratefully acknowledge NVIDIA corporation for the donation of Tesla GPUs for this research.  ... 
arXiv:1411.6067v2 fatcat:5vrafvdy3zfnjde7iurkpm7upq

Object recognition and segmentation by non-rigid quasi-dense matching

Juho Kannala, Esa Rahtu, Sami S. Brandt, Janne Heikkila
2008 2008 IEEE Conference on Computer Vision and Pattern Recognition  
The matching method is based on the match propagation algorithm which is here extended by using local image gradients for adapting the propagation to smooth non-rigid deformations of the imaged surfaces  ...  Our approach for object recognition and segmentation is directly built on the quasi-dense matching.  ...  Aknowledgements We are grateful to Vittorio Ferrari for providing his data for the experiments. We would also like to thank the anonymous reviewers for their feedback.  ... 
doi:10.1109/cvpr.2008.4587472 dblp:conf/cvpr/KannalaRBH08 fatcat:i5boieeahvgnrj77vfcuyh7hpa

3D object class detection in the wild

Bojan Pepik, Michael Stark, Peter Gehler, Tobias Ritschel, Bernt Schiele
2015 2015 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)  
Our 3D object class detection method consists of several stages gradually enriching the object detection output with object viewpoint, keypoints and 3D shape estimates.  ...  Object class detection has been a synonym for 2D bounding box localization for the longest time, fueled by the success of powerful statistical learning techniques, combined with robust image representations  ...  regression, (iii) object keypoint detection and (iv) 3D lifting through rigid keypoint alignment.  ... 
doi:10.1109/cvprw.2015.7301358 dblp:conf/cvpr/Pepik0GRS15 fatcat:22qhixirr5fkrgcz46joouwvry
« Previous Showing results 1 — 15 out of 23,910 results