Filters








10 Hits in 1.1 sec

ClassCut for Unsupervised Class Segmentation [chapter]

Bogdan Alexe, Thomas Deselaers, Vittorio Ferrari
2010 Lecture Notes in Computer Science  
We propose a novel method for unsupervised class segmentation on a set of images. It alternates between segmenting object instances and learning a class model.  ...  image independently; (b) outperforms Grabcut [1] for the task of unsupervised segmentation; (c) offers competitive performance compared to the state-of-the-art in unsupervised segmentation and in particular  ...  Conclusion We presented a novel approach to unsupervised class segmentation.  ... 
doi:10.1007/978-3-642-15555-0_28 fatcat:tcwul3bjsbcgbnoy6c2z6tx6hm

A hierarchical graph model for object cosegmentation

Yanli Li, Zhong Zhou, Wei Wu
2013 EURASIP Journal on Image and Video Processing  
The results verify that our method achieves better segmentation quality as well as higher efficiency.  ...  Besides, a histogram based saliency detection scheme is employed for initialization. We demonstrate experimental evaluations with state-of-the-art methods over several public datasets.  ...  Another technique, also called unsupervised object segmentation such as LOCUS [8] , ClassCut [9] , Arora et al. [10] and Chen et al.  ... 
doi:10.1186/1687-5281-2013-11 fatcat:prhxngr3m5flze2u7m4e6p6rqu

Learning what is where from unlabeled images: joint localization and clustering of foreground objects

Ashok Chandrashekar, Lorenzo Torresani, Richard Granger
2013 Machine Learning  
We describe two methods for efficient foreground localization: the first method does not require any bottom-up image segmentation and discovers the foreground region as a contiguous rectangular bounding  ...  However, unlike previous methods, objects are not assumed to be encapsulated by a single segment.  ...  Acknowledgements The authors world like to thank Minh Hoai Nguyen for sharing his code for cosegmentation of image pairs.  ... 
doi:10.1007/s10994-013-5330-2 fatcat:kit34ajhqvdzffiopz7wxcsolq

Factored Shapes and Appearances for Parts-based Object Understanding

Seyed Mohammadali Eslami, Christopher Williams
2011 Procedings of the British Machine Vision Conference 2011  
We present a novel generative framework for learning parts-based representations of object classes.  ...  We propose Markov Chain Monte Carlo sampling schemes for efficient inference and learning, and evaluate the model on a number of datasets.  ...  We are grateful for financial support from the Carnegie Trust and the SORSAS scheme for AE.  ... 
doi:10.5244/c.25.18 dblp:conf/bmvc/EslamiW11 fatcat:xpypyeqj3vgbhhbnimm6d2r4hu

Distilled Collections from Textual Image Queries

Hadar Averbuch-Elor, Yunhai Wang, Yiming Qian, Minglun Gong, Johannes Kopf, Hao Zhang, Daniel Cohen-Or
2015 Computer graphics forum (Print)  
Our approach is unsupervised, built on a novel clustering scheme, and solves the distillation and object segmentation problems simultaneously.  ...  In addition, the object of interest is properly segmented out throughout the distilled set.  ...  The ClassCut [ADF10] technique, for example, aims at cosegmenting a set of images capturing object instances of an unknown class.  ... 
doi:10.1111/cgf.12547 fatcat:mpypdujyzrhahos7oy5prisgum

Which Image Pairs Will Cosegment Well? Predicting Partners for Cosegmentation [chapter]

Suyog Dutt Jain, Kristen Grauman
2015 Lecture Notes in Computer Science  
This is problematic, since coupling the "wrong" images for segmentationeven images of the same object class-can actually deteriorate performance relative to single-image segmentation.  ...  Cosegmentation methods segment multiple related images jointly, exploiting their shared appearance to generate more robust foreground models.  ...  We show the 10 best and 4 worst performing classes (see Supp. for all classes).  ... 
doi:10.1007/978-3-319-16811-1_12 fatcat:woqvn2po7zaoxort272qq37jyy

Integrating low-level and semantic features for object consistent segmentation

Hao Fu, Guoping Qiu
2013 Neurocomputing  
Numerous methods for semantic segmentation have been proposed in recent years and most of them chose pixel or superpixel as the processing primitives.  ...  The aim of semantic segmentation is to assign each pixel a semantic label.  ...  For example, there are only about 10 regions for some semantic class in MSRC21 dataset. Therefore it is very easy for the classifier to get overfitted.  ... 
doi:10.1016/j.neucom.2012.01.050 fatcat:ztnx2by2zbau3ki5spapkgmiqq

Integrating Low-level and Semantic Features for Object Consistent Segmentation

Hao Fu, Guoping Qiu
2011 2011 Sixth International Conference on Image and Graphics  
Numerous methods for semantic segmentation have been proposed in recent years and most of them chose pixel or superpixel as the processing primitives.  ...  The aim of semantic segmentation is to assign each pixel a semantic label.  ...  For example, there are only about 10 regions for some semantic class in MSRC21 dataset. Therefore it is very easy for the classifier to get overfitted.  ... 
doi:10.1109/icig.2011.47 dblp:conf/icig/FuQ11 fatcat:znzfqzo42fehpoac44bfxohcyi

SEMANTIC SEGMENTATION USING GRABCUT
english

2012 Proceedings of the International Conference on Computer Vision Theory and Applications   unpublished
C-GrabCut generates multiple class-specific segmentations and classifies them by using shape and color information.  ...  This work analyzes how to utilize the power of the popular GrabCut algorithm for the task of pixel-wise labeling of images, which is also known as semantic segmentation and an important step for scene  ...  Therefore, we have a segmentation of an image for each class.  ... 
doi:10.5220/0003829905970602 fatcat:kcuonw5ylvfejgxjpt7i4g6udq

Human machine collaboration for foreground segmentation in images and videos [article]

Suyog Dutt Jain
2018
Interactive image segmentation with active human input 9 1.1.2 Interactive image and video segmentation with point clicks 1.1.3 Active segmentation propagation in image collections . .  ...  Unsupervised video segmentation Fully automatic or unsupervised video segmentation methods assume no human input on the video.  ...  Two existing unsupervised methods also incorporate the Robust P n model for video segmentation, but with important differences from my approach.  ... 
doi:10.15781/t2r20sc9c fatcat:zqw2v4snbvdnvhb4v42cx7xk4e