Filters








9 Hits in 2.5 sec

BoLTVOS: Box-Level Tracking for Video Object Segmentation [article]

Paul Voigtlaender and Jonathon Luiten and Bastian Leibe
2019 arXiv   pre-print
We approach video object segmentation (VOS) by splitting the task into two sub-tasks: bounding box level tracking, followed by bounding box segmentation.  ...  Following this paradigm, we present BoLTVOS (Box-Level Tracking for VOS), which consists of an R-CNN detector conditioned on the first-frame bounding box to detect the object of interest, a temporal consistency  ...  We would like to thank Bo Li for helpful discussions.  ... 
arXiv:1904.04552v2 fatcat:tgf74nhxkvclpkdbv2lbgfkrpu

Exploring the Combination of PReMVOS, BoLTVOS and UnOVOST for the 2019 YouTube-VOS Challenge

Jonathon Luiten, Paul Voigtlaender, Bastian Leibe
2019 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)  
Video Object Segmentation is the task of tracking and segmenting objects in a video given the first-frame mask of objects to be tracked.  ...  There have been a number of different successful paradigms for tackling this task, from creating object proposals and linking them in time as in PRe-MVOS, to detecting objects to be tracked conditioned  ...  These are PReMVOS [10] (Proposal-generation, Refinement and Merging for VOS), BoLTVOS [16] (Box-Level Tracking for VOS), and Un-OVOST [19] (Unsupervised Offline VOS and Tracking). PReMVOS.  ... 
doi:10.1109/iccvw.2019.00087 dblp:conf/iccvw/LuitenVL19 fatcat:sf53s33mhvhspbnm77hcgcnza4

Query-Memory Re-Aggregation for Weakly-supervised Video Object Segmentation

Fanchao Lin, Hongtao Xie, Yan Li, Yongdong Zhang
2021 AAAI Conference on Artificial Intelligence  
Weakly-supervised video object segmentation (WVOS) is an emerging video task that can track and segment the target given a simple bounding box label.  ...  solve the problem, we propose a novel Re-Aggregation based framework, which uses feature matching to efficiently find the target and capture the temporal dependencies from multiple frames to guide the segmentation  ...  BoltVOS (Voigtlaender, Luiten, and Leibe 2019) splits the WVOS task into two sub-tasks: the box-level tracking, and the segmentation of the bounding box.  ... 
dblp:conf/aaai/LinX0021 fatcat:xpgt2meivng7pd5wzi5k4zsdcm

Fast Template Matching and Update for Video Object Tracking and Segmentation

Mingjie Sun, Jimin Xiao, Eng Gee Lim, Bingfeng Zhang, Yao Zhao
2020 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
In this paper, the main task we aim to tackle is the multiinstance semi-supervised video object segmentation across a sequence of frames where only the first-frame box-level ground-truth is provided.  ...  To sum up, most video object tracking and segmentation algorithms consist of three steps.  ...  To overcome this problem, inspired by the rapid progress in the task of video object tracking (VOT) at bounding box level, some works attempt to rely on the firstframe bounding boxes to provide target  ... 
doi:10.1109/cvpr42600.2020.01080 dblp:conf/cvpr/SunXLZ020 fatcat:ymsce5ph45dmdcphspyvoeaubm

Fast Template Matching and Update for Video Object Tracking and Segmentation [article]

Mingjie Sun, Jimin Xiao, Eng Gee Lim, Bingfeng Zhang, Yao Zhao
2020 arXiv   pre-print
In this paper, the main task we aim to tackle is the multi-instance semi-supervised video object segmentation across a sequence of frames where only the first-frame box-level ground-truth is provided.  ...  To sum up, most video object tracking and segmentation algorithms consist of three steps.  ...  To overcome this problem, inspired by the rapid progress in the task of video object tracking (VOT) at bounding box level, some works attempt to rely on the firstframe bounding boxes to provide target  ... 
arXiv:2004.07538v1 fatcat:zq6vevbrsrdaflewrkimorktxy

Efficient Regional Memory Network for Video Object Segmentation [article]

Haozhe Xie, Hongxun Yao, Shangchen Zhou, Shengping Zhang, Wenxiu Sun
2021 arXiv   pre-print
Recently, several Space-Time Memory based networks have shown that the object cues (e.g. video frames as well as the segmented object masks) from the past frames are useful for segmenting objects in the  ...  For the current query frame, the query regions are tracked and predicted based on the optical flow estimated from the previous frame.  ...  Lucid-Tracker [14] synthesizes in-domain data to train a specialized pixel-level video object segmenter.  ... 
arXiv:2103.12934v2 fatcat:bvakp4um7zewhlujrri36kugoq

Contextual Guided Segmentation Framework for Semi-supervised Video Instance Segmentation [article]

Trung-Nghia Le and Tam V. Nguyen and Minh-Triet Tran
2021 arXiv   pre-print
In this paper, we propose Contextual Guided Segmentation (CGS) framework for video instance segmentation in three passes.  ...  For human instance, we develop skeleton-guided segmentation in a frame along with object flow to correct and refine the result across frames.  ...  We also thank NVIDIA and AIOZ Pte Ltd for the support of GPU and computing infrastructure.  ... 
arXiv:2106.03330v1 fatcat:2gku7u36lne3dp3zvauquexwjq

An Exploration of Target-Conditioned Segmentation Methods for Visual Object Trackers [article]

Matteo Dunnhofer, Niki Martinel, Christian Micheloni
2020 arXiv   pre-print
Visual object tracking is the problem of predicting a target object's state in a video.  ...  vision community, in order to transform any bounding-box tracker into a segmentation tracker.  ...  From a more general point of view, the recent video object segmentation (VOS) problem requires to produce the segmentation masks of generic target objects in a video, given the mask of each in the first  ... 
arXiv:2008.00992v2 fatcat:tygm4fhuefho7itzxhpez6ieiu

Table of contents

2019 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)  
), Yunchao Wei (University of Technology Sydney, Australia), and Yi Yang (University of Technology Sydney, Australia) Towards Good Practices for Video Object Segmentation 701 Dongdong Yu (ByteDance  ...  University), Peisen Wang (Megvii Research), Haoqiang Fan (Megvii Research), and Si Liu (Beihang University) Motion-Guided Spatial Time Attention for Video Object Segmentation 693 Qiang Zhou (Huazhong  ... 
doi:10.1109/iccvw.2019.00004 fatcat:balgnbs6n5gbvosrkz3mx5gn6m