Towards automatic discovery of object categories

By Markus Weber, Max Welling, Pietro Perona. Proceedings IEEE Conference on Computer Vision and Pattern Recognition. CVPR 2000 (Cat. No.PR00662), 2000

325 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/472ff211d18aa2ca2b65b69d86499430ad287499

How researchers see this paper

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

In Computer Vision and Pattern Recognition

Citations in the corpus

Learning methods for generic object recognition with invariance to pose and lighting (2004, background): “Authors have advocated the use of color, texture, and contours for image indexing applications [8], the detection of distinctive local features [20] [26] [25] [23], the use of global appearance templates [11] [10] [19], the extraction of silhouettes and edge information [14, 22, 8, 4, 19] and the…”
Local Features and Kernels for Classification of Texture and Object Categories: A Comprehensive Study (2006, methodology): “The Xerox7 dataset (Willamowski et al., 2004) consists of 1776 images of seven classes: bikes, books, buildings, cars, faces, phones and trees. This is a challenging dataset, as it includes images with highly variable pose and background clutter, and the intra-class variability is large. Some of the”
Large-scale Learning with SVM and Convolutional for Generic Object Categorization (2006, background): “The use of color, texture, and contours have been advocated in [16], the use of global appearance templates in [19, 18, 25], silhouettes and edge information [22, 29, 16, 5, 25], pose-invariant feature histograms [17, 6, 1], distinctive local features [26, 35, 33, 12], and image fragments [30].”
Learning part-based templates from large collections of 3D shapes (2013, methodology): “In contrast, we learn the part-based model from unlabeled data, evolving a set of templates to best fit the data (see also [Weber et al. 2000a; Weber et al. 2000b] for applications in visual recognition).”
Unsupervised Object Discovery: A Comparison (2010, background): “1.1 Related Work Probably the first work to tackle the problem of unsupervised object category discovery has been Weber et al. (2000), building on a simplified constellation model-like scheme.”
LabelMe: Online Image Annotation and Applications (2010, unknown): “D”
Learning Active Basis Model for Object Detection and Recognition (2010, background): “Figure 60 The recursive active basis can be considered a constellation model (Weber et al. 2000) whose constituent components are active bases.”
Incremental learning of nonparametric Bayesian mixture models (2008, background): “Discovering visual categories automatically with minimal human supervision is perhaps the most exciting current challenge in machine vision [21, 16].”
I2T: Image Parsing to Text Description (2010, methodology): “This is equivalent to other part-based models such as the constellation model [53] and the pictorial structures [54].”
View-Invariant Recognition Using Corresponding Object Fragments (2004, methodology): “This can be contrasted with approaches such as [ 22 ], where the features are constrained to have a fixed size.”
Object Detection with Discriminatively Trained Part Based Models (2010, background): “One popular approach for part-based models is to restrict part locations to a small set of possible locations returned by an interest point detector [1], [18], [42].”
Unsupervised Learning of Object Landmarks by Factorized Spatial Embeddings (2017, methodology): “Several unsupervised methods to learn parts exist, from the constellation approach used in [19, 9, 62] to the Deformable Parts Model (DPM) [18] and many others.”
Robust Object Detection with Interleaved Categorization and Segmentation (2008, methodology): “Many approaches employ k-means clustering because of its computational simplicity, which allows to apply it to very large data sets (Weber et al. 2000).”
Stacked Autoencoders for Unsupervised Feature Learning and Multiple Organ Detection in a Pilot Study Using 4D Patient Data (2013, methodology): “So-called " partmodels " for the self-learning of object classes were studied for 2D images in [11], [12], [13], [14], and [15] to achieve object detection in such weakly supervised settings.”
Blocks That Shout: Distinctive Parts for Scene Classification (2013, unknown): “The notion of part has been of central importance in object recognition since the introduction of pictorial structures [11], constellation models [35], object fragments [2, 29], right up to recent state-of-the-art methods such as Deformable Part Models (DPMs) [9].”
Large scale mining and retrieval of visual data in a multimodal context (2008, methodology): “even localize instances of object classes in images [Weber et al., 2000a; Agarwal and Roth, 2002]”
Unsupervised Learning of Individuals and Categories from Images (2008, background): “Weber, Welling, and Perona (2000) also cast the unsupervised categorization problem as emergent population coding, but again present images from only a single category at a time.”