Filters








169,386 Hits in 4.2 sec

Consistent depth of moving objects in video

Zhoutong Zhang, Forrester Cole, Richard Tucker, William T. Freeman, Tali Dekel
2021 ACM Transactions on Graphics  
We present a method to estimate depth of a dynamic scene, containing arbitrary moving objects, from an ordinary video captured with a moving camera.  ...  Our depth maps give rise to a number of depth-and-motion aware video editing effects such as object and lighting insertion.  ...  Consistent Depth of Moving Objects in Video • 148:7 ACM Trans. Graph., Vol. 40, No. 4, Article 148. Publication date: August 2021.  ... 
doi:10.1145/3450626.3459871 fatcat:3syvszwnl5cwhbaqkphv63vdca

Refilming with Depth-Inferred Videos

Guofeng Zhang, Zilong Dong, Jiaya Jia, Liang Wan, Tien-Tsin Wong, Hujun Bao
2009 IEEE Transactions on Visualization and Computer Graphics  
These tools can be utilized to produce a variety of visual effects in our system, including but not limited to video composition, "predator" effect, bullet-time, depth-of-field, and fog synthesis.  ...  In this paper, we present a new video editing system for creating spatiotemporally consistent and visually appealing refilming effects.  ...  They also thank Hanqing Jiang, Li Xu, Jianing Chen, Hin Shun Chung, Liansheng Wang, and Xiaopei Liu for their help in user testing, video capture, and production, and Carl Jantzen for video narration.  ... 
doi:10.1109/tvcg.2009.47 pmid:19590108 fatcat:pd64areexjgb5irrpzoqgr45c4

Learning Depth from Monocular Videos Using Synthetic Data: A Temporally-Consistent Domain Adaptation Approach [article]

Yipeng Mou, Mingming Gong, Huan Fu, Kayhan Batmanghelich, Kun Zhang, Dacheng Tao
2019 arXiv   pre-print
constraints in the videos to improve style transfer and depth prediction.  ...  In this paper, we propose to resolve this dilemma by transferring knowledge from synthetic videos with easily obtainable ground-truth depth labels.  ...  Furthermore, given that the temporal consistency in real videos guided by camera pose and depth is sensitive to the moving objects in the scene, we further proposed a moving mask prediction network trained  ... 
arXiv:1907.06882v2 fatcat:aakfbukbn5gvdhlio7prc2u6ai

Unsupervised Monocular Depth Learning in Dynamic Scenes [article]

Hanhan Li, Ariel Gordon, Hang Zhao, Vincent Casser, Anelia Angelova
2020 arXiv   pre-print
We present a method for jointly training the estimation of depth, ego-motion, and a dense 3D translation field of objects relative to the scene, with monocular photometric consistency being the sole source  ...  , and they tend to be constant for rigid moving objects.  ...  [12, 41] estimate the motion of objects in the scenes, with the assistance of pre-trained segmentation models, leading to significant improvement in depth estimation for moving objects.  ... 
arXiv:2010.16404v2 fatcat:dwf7hypltnbijn3somdt6hv2zu

Instance-wise Depth and Motion Learning from Monocular Videos [article]

Seokju Lee, Sunghoon Im, Stephen Lin, In So Kweon
2020 arXiv   pre-print
Second, we design an instance-wise photometric and geometric consistency loss that effectively decomposes background and moving object regions.  ...  We present an end-to-end joint training framework that explicitly models 6-DoF motion of multiple dynamic objects, ego-motion and depth in a monocular camera setup without supervision.  ...  Given two consecutive frames from a video, the proposed neural network produces depth, 6-DoF motion of each moving object, and the ego-motion between adjacent frames as shown in Fig. 1 .  ... 
arXiv:1912.09351v2 fatcat:knjqipriwfbldescp36mtp3buq

Moving Object Extraction with a Hand-held Camera

Guofeng Zhang, Jiaya Jia, Wei Xiong, Tien-Tsin Wong, Pheng-Ann Heng, Hujun Bao
2007 2007 IEEE 11th International Conference on Computer Vision  
Experimental results of high quality moving object extraction from challenging videos demonstrate the effectiveness of our method.  ...  In our method, based on the robust motion estimation, we are capable of handling challenging videos where the background contains complex depth and the camera undergoes unknown motions.  ...  For instance, if in a video capturing, the camera undergoes the same motion as the foreground object in order to keep it in center of all frames, the computed residual errors of the pixels on moving object  ... 
doi:10.1109/iccv.2007.4408963 dblp:conf/iccv/ZhangJXWHB07 fatcat:tjcqe4snevbmjcxnhwxve2tyfi

Fast Generation Algorithm of Digital Hologram based Depth Difference Temporal Filtering

Hyun-Jun Choi
2013 International Journal of Multimedia and Ubiquitous Engineering  
This paper proposes an algorithm that increases the speed of generating a digital hologram using a depth different temporal filtering (DDTF) of a depth video sequence.  ...  Thus it has been developed to get some ideal characteristics of CGH that are not possible in reality or to test the characteristics of a hologram.  ...  Depth-Map Video of a Horizontally Moving Object Figure 8.  ... 
doi:10.14257/ijmue.2013.8.6.17 fatcat:q5lfk5sl2rbb3kkj5t5iadl77i

Positional Information is All You Need: A Novel Pipeline for Self-Supervised SVDE from Videos [article]

Juan Luis Gonzalez Bello, Jaeho Moon, Munchurl Kim
2022 arXiv   pre-print
Finally, we employ existing boosting techniques in a new way to further self-supervise the depth of the moving objects.  ...  in the scenes, allowing for better learning of SVDE from videos.  ...  Finally, we employ existing boosting techniques in a new way to further self-supervise the depth of the moving objects.  ... 
arXiv:2205.08851v1 fatcat:x5iajs3vsjhxzcwd4jd2b7jwsq

Consistent Video Depth Estimation [article]

Xuan Luo, Jia-Bin Huang, Richard Szeliski, Kevin Matzen, Johannes Kopf
2020 arXiv   pre-print
We present an algorithm for reconstructing dense, geometrically consistent depth for all pixels in a monocular video.  ...  At test time, we fine-tune this network to satisfy the geometric constraints of a particular input video, while retaining its ability to synthesize plausible depth details in parts of the video that are  ...  ACKNOWLEDGMENTS We would like to thank Patricio Gonzales Vivo, Dionisio Blanco, and Ocean Quigley for creating the artistic effects in the accompanying video.  ... 
arXiv:2004.15021v2 fatcat:k67ia5ez4rdzxc6phce4uo4hhy

SfM-Net: Learning of Structure and Motion from Video [article]

Sudheendra Vijayanarasimhan, Susanna Ricco, Cordelia Schmid, Rahul Sukthankar, Katerina Fragkiadaki
2017 arXiv   pre-print
We propose SfM-Net, a geometry-aware neural network for motion estimation in videos that decomposes frame-to-frame pixel motion in terms of scene and object depth, camera motion and 3D object rotations  ...  It often successfully segments the moving objects in the scene, even though such supervision is never provided.  ...  This shows that it is important to account for moving objects when training on videos in the wild.  ... 
arXiv:1704.07804v1 fatcat:e3qxconotfgajazkuxjnz5yyp4

Learning the Depths of Moving People by Watching Frozen People [article]

Zhengqi Li, Tali Dekel, Forrester Cole, Richard Tucker, Noah Snavely, Ce Liu, William T. Freeman
2019 arXiv   pre-print
In this paper, we take a data-driven approach and learn human depth priors from a new source of data: thousands of Internet videos of people imitating mannequins, i.e., freezing in diverse, natural poses  ...  Existing methods for recovering depth for dynamic, non-rigid objects from monocular video impose strong assumptions on the objects' motion and may only recover sparse depth.  ...  In summary, our contributions are: i) a new source of data for depth prediction consisting of a large number of Internet videos in which the camera moves around people "frozen" in natural poses, along  ... 
arXiv:1904.11111v1 fatcat:b2zvj2rdlzb6hkjuqzijtiqcgy

Learning the Depths of Moving People by Watching Frozen People

Zhengqi Li, Tali Dekel, Forrester Cole, Richard Tucker, Noah Snavely, Ce Liu, William T. Freeman
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
In this paper, we take a data-driven approach and learn human depth priors from a new source of data: thousands of Internet videos of people imitating mannequins, i.e., freezing in diverse, natural poses  ...  Existing methods for recovering depth for dynamic, non-rigid objects from monocular video impose strong assumptions on the objects' motion and may only recover sparse depth.  ...  In summary, our contributions are: i) a new source of data for depth prediction consisting of a large number of Internet videos in which the camera moves around people "frozen" in natural poses, along  ... 
doi:10.1109/cvpr.2019.00465 dblp:conf/cvpr/LiDCTSLF19 fatcat:p6qcmicraffr7esiuksvduxxbe

Revisiting Depth Layers from Occlusions

Adarsh Kowdle, Andrew Gallagher, Tsuhan Chen
2013 2013 IEEE Conference on Computer Vision and Pattern Recognition  
In this work, we consider images of a scene with a moving object captured by a static camera.  ...  prior for the moving object.  ...  In case of k moving objects, we add k nodes (a node for each moving object) for each frame of the video. The resulting spatio-temporal graph for the example is shown in Figure 5(b) .  ... 
doi:10.1109/cvpr.2013.272 dblp:conf/cvpr/KowdleGC13 fatcat:mxzvrmpay5gxtcugytjhir6zsi

Geo-spatial aerial video processing for scene understanding and object tracking

Jiangjian Xiao, Hui Cheng, Feng Han, Harpreet Sawhney
2008 2008 IEEE Conference on Computer Vision and Pattern Recognition  
coordinate system for pixels in the video.  ...  labeling of structures such as buildings, foliage, and roads for scene understanding, and (2) elimination of moving object detection and tracking errors using 3D parallax constraints and semantic labels  ...  as well as moving objects.  ... 
doi:10.1109/cvpr.2008.4587434 dblp:conf/cvpr/XiaoCHS08 fatcat:5ejclvnxbzhypdro776px3ovma

Region Deformer Networks for Unsupervised Depth Estimation from Unconstrained Monocular Videos [article]

Haofei Xu, Jianmin Zheng, Jianfei Cai, Juyong Zhang
2019 arXiv   pre-print
The core contribution lies in RDN for proper handling of rigid and non-rigid motions of various objects such as rigidly moving cars and deformable humans.  ...  In this paper, we propose a new learning based method consisting of DepthNet, PoseNet and Region Deformer Networks (RDN) to estimate depth from unconstrained monocular videos without ground truth supervision  ...  Zou et al. [2018] jointly learn depth and optical flow from monocular videos with a cross-task consistency loss in the rigid scene, but the depth of moving objects does not benefit from the learned optical  ... 
arXiv:1902.09907v2 fatcat:4jfeqpptanghnhkjkq5hojrod4
« Previous Showing results 1 — 15 out of 169,386 results