LOCUS: learning object classes with unsupervised segmentation

By John M. Winn, Nebojsa Jojic. Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1, 2005

581 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/79bd5e61e20594c13a90e66293e1abb4179935a2

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

TextonBoost: Joint Appearance, Shape and Context Modeling for Multi-class Object Recognition and Segmentation (2006, background): “Joint detection and segmentation of a single object class has been achieved by several authors [7, 8, 9].”
The Layout Consistent Random Field for Recognizing and Segmenting Partially Occluded Objects (2006, unknown): “More recently, Winn and Jojic [18] used a dense generative model to learn a partitioning of the object into parts, along with an unsupervised segmentation of the object.”
Learning object class detectors from weakly annotated video (2012, background): “Because manual annotation can be inflexible, expensive and tedious, recent work investigated methods that can learn from unlabeled or weakly annotated data [3, 6, 9, 23, 37].”
Beyond pixels: exploring new representations and applications for motion analysis (2009, background): “Sophisticated object representations [10, 12, 34, 129] have been developed to cope with the variations in objects' shape and appearance.”
Multi-View K-Means Clustering on Big Data (2013, unknown): “, 2007], Microsoft Research Cambridge Volume 1(MSRC-v1) [Winn and Jojic, 2005] Handwritten numerals [Frank and Asuncion, 2010], Animal with attribute [Lampert et al.”
Using Segmentation to Verify Object Hypotheses (2007, methodology): “This simple approach is similar to the unsupervised algorithm from [17], though it avoids an explicit registration step by leveraging ground-truth bounding boxes.”
What is an object? (2010, background): “It can facilitate learning new classes in a weakly supervised scenario [6], where the location of object instances is unknown [12, 34].”
Weakly supervised semantic segmentation with a multi-image model (2011, background): “Also related are single-class weakly supervised segmentation methods [1, 11].”
LabelMe: Online Image Annotation and Applications (2010, background): “A complementary line of research tries to avoid the need to annotate images by developing unsupervised learnin g algorithms [13], [65], [81], [80], [15], [59], [49], [68], [ 16].”
Learning to detect unseen object classes by between-class attribute transfer (2009, methodology): “Similarly, image segmentation has been proposed as an unsupervised method to extract contours that are discriminative for object classes [37].”
Parameter-Free Auto-Weighted Multiple Graph Learning: A Framework for Multiview Clustering and Semi-Supervised Classification (2016, methodology): “We evaluate the performance of the proposed framework AMGL for clustering and semi-supervised classification on the following four data sets: MSRC-v1 data set [Winn and Jojic, 2005] contains 240 images and can be divided into 8 classes.”
Multi-view Subspace Clustering (2015, background): “That is Caltech-101 [9], Microsoft Research Cambridge Volume [23] and the ETH Zurich ETH-80 [15] .”
Modeling object appearance using Context-Conditioned Component Analysis (2015, background): “Winn and Jojic introduced LOCUS [30], an unsupervised generative model that learns segmentations of visual objects.”
Object cosegmentation (2011, methodology): “Although [24] reports significantly better error rates than [10], it uses shape features, assuming a rough alignment of all the objects, given a reference frame.”
Delving deeper into the whorl of flower segmentation (2010, background): “The LOCUS approach of Winn and Jojic [1] is a good example, where a shape matte (e.”
Learning multiple layers of representation. (2007, background): “Learning by optimizing a variational bound is now a standard way of dealing with the intractability of inference in complex generative models[24, 25, 26, 27].”
Symbiotic Segmentation and Part Localization for Fine-Grained Categorization (2013, background): “A representative early work in this area is LOCUS [31].”
Discriminative clustering for image co-segmentation (2010, background): “At the same time, in the context of object recognition, where object instances may vary in pose, shape or color, co-segmentation should provide mid-level features which could improve recognition performance, [16, 20, 23].”