Conditional Random Fields for Object Recognition

By Ariadna Quattoni, Michael Collins, Trevor Darrell. Neural Information Processing Systems, 2004

439 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/819a2bbd4467192dbbc9e833eb4ca6653ad791aa

How researchers see this paper

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

Citations in the corpus

Learning to Combine Bottom-Up and Top-Down Segmentation (2006, methodology): “The CRF formulation has proven useful in many vision applica tions [7, 15, 14, 4, 5].”
A Comprehensive Survey of Scene Graphs: Generation and Application (2021, unknown): “Similar CRFs are widely utilized in computer vision tasks [37], [40] and have been demonstrated effective in capturing statistical correlations in visual relationships.”
FACTORIE: Probabilistic Programming via Imperatively Defined Factor Graphs (2009, background): “Conditional random fields [1], or discriminatively trained undirected graphical models, have become the tool of choice for addressing many important tasks across bioinformatics, natural language processing, robotics, and many other fields [2, 3, 4].”
Hidden Conditional Random Fields (2007, background): “In parallel to our work on object recognition [18], [6] developed a hidden-state CRF model for phone recognition.”
Structured Models for Fine-to-Coarse Sentiment Analysis (2007, methodology): “If viewed this way, it is also possible to run a constrained forward-backward algorithm and learn the parameters for CRFs as well.”
Using Combination of Statistical Models and Multilevel Structural Information for Detecting Urban Areas From a Single Gray-Level Image (2007, methodology): “In the later research, the CRFs have been generalized to image labeling, such as image segmentation [14], image classification [15]–[17], and object recognition [18], [19].”
Extracting Places and Activities from GPS Traces Using Hierarchical Conditional Random Fields (2007, background): “However, discriminative models such as conditional random fields (CRF), have recently been shown to outperform generative techniques in areas such as natural language processing [18, 37], information extraction [31, 15], web page classification [38], and computer vision [17, 33].”
Latent-Dynamic Discriminative Models for Continuous Gesture Recognition (2007, methodology): “In the vision community, HCRFs have been used to model spatial dependencies for object recognition in cluttered images [18] and for arm and head gesture recognition from segmented sequences [27].”
Conditional random fields for activity recognition (2007, methodology): “As newer models, conditional random fields have received less attention than HMMs or DBNs, but they are used for activity recognition as well as in related domains that require rich, continuous features such as image segmentation [54], object recognition [88], and gesture recognition [116].”
Towards Holistic Scene Understanding: Feedback Enabled Cascaded Classification Models (2010, methodology): “hidden conditional rand om field [29], latent structured SVM [30]), which can be potentially applied to multi-task settings with disjoint datase ts.”
Continuous body and hand gesture recognition for natural human-computer interaction (2012, methodology): “Our approach computes a weighted mean of neighboring input features, which does not increase the dimensionality or complexity compared to the original HCRF model [Quattoni et al. 2005] (additions and multiplications in the kernel operation can be negligible compared to the complexity of the inferenc”
The Layout Consistent Random Field for Recognizing and Segmenting Partially Occluded Objects (2006, methodology): “A number of methods use a Conditional Random Field (CRF) [10] to achieve segmentation [9] or sparse part-based detection [14].”
Iterative parameter mixing for distributed large-margin training of structured predictors for natural language processing (2015, background): “For example, one can straightforwardly use latent variables in an effort to maximize conditional likelihood (Quattoni et al., 2004; Koo & Collins, 2005; Petrov & Klein, 2007a).”
Supervised Learning of Image Restoration with Convolutional Networks (2007, background): “Some have argued that CRFs are superior to MRFs for image analysis [11, 13, 23].”
Scene Segmentation with CRFs Learned from Partially Labeled Images (2007, methodology): “CRF models can be applied either at the pixel-level [5, 6, 7] or at the coarser level of super-pixels or patches [8, 9, 10].”