Semantic texton forests for image categorization and segmentation

By Jamie Shotton, Matthew Johnson, Roberto Cipolla. 2008

1,156 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

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 to Refine Object Segments (2016, background): “The task is related to both object detection with bounding boxes [9–11] and semantic segmentation [10, 12–19].”
DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs (2016, background): “Most of the successful semantic segmentation systems developed in the previous decade relied on hand-crafted features combined with flat classifiers, such as Boosting [24], [42], Random Forests [43], or Support Vector Machines [44].”
Deep Interactive Object Selection (2016, methodology): “Many algorithms have been proposed such as superpixel-based [28, 24, 25] and clusteringbased methods [23, 15, 14].”
Conditional Random Fields as Recurrent Neural Networks (2015, methodology): “Works on this topic include: TextonBoost [28], TextonForest [27], and Random Forest-based human body pose estimation for Kinect [26].”
SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation (2015, methodology): “Another approach focuses on real-time joint reconstruction and semantic segmentation, where Random Forests are used as the classifier [35].”
Unsupervised Model Adaptation for Continual Semantic Segmentation (2020, methodology): “Traditional semantic segmentation algorithms use handengineered extracted features which are fed into a classifier (Shotton, Johnson, and Cipolla 2008; Tighe and Lazebnik 2010), where the classifier is trained using supervised learning.”
Knowledge Distillation for Incremental Learning in Semantic Segmentation (2019, background): “dataset creators [50, 57], because they are underrepresented inside the dataset.”
From image-level to pixel-level labeling with Convolutional Networks (2014, methodology): “To circumvent this issue, we consider simple post-processing techniques, namely image-level prior (ILP) and three different smoothing priors (SP), with increasing amount of information .”
Full-Resolution Residual Networks for Semantic Segmentation in Street Scenes (2016, background): “Semantic image segmentation [49, 38, 21, 53, 33], the task of assigning a set of predefined class labels to image pixels, is an important tool for modeling the complex relationships of the semantic entities usually found in street scenes, such as cars, pedestrians, road, or sidewalks.”
Decision tree fields (2011, methodology): “Since for most computer vision applications it is well known that the key to obtaining high predictive performance is the amount of training data, many recent works use decision trees, or a related variant of it (random Forests [6], extremely randomized trees [10], semantic texton forest [28]).”
Real-time Action Recognition by Spatiotemporal Semantic and Structural Forests (2010, methodology): “The recognition accuracy is enhanced by adaptively combining PSRM and the bag of semantic texton (BOST) method [6]: the k-means forest classifier is learned using PSRM as a matching kernel.”
Deep semantic segmentation for automated driving: Taxonomy, roadmap and challenges (2017, methodology): “In [17] decision 2017 IEEE 20th International Conference on Intelligent Transportation Systems (ITSC)”
Transition Forests: Learning Discriminative Temporal Transitions for Action Recognition and Detection (2016, background): “A related line of work [13, 18, 21, 23] proposes decision forests methods for image segmentation.”
On-line Random Forests (2009, background): “RFs have demonstrated to be better or at least comparable to other stateof-the-art methods in both classification [6] [4], semantic segmentation [21], real-time keypoint recognition [13] and clustering applications [14].”
Structured class-labels in random forests for semantic image labelling (2011, neutral): “Texton forests naı̈ve (supervised) [26] 49.”
ROAD: Reality Oriented Adaptation for Semantic Segmentation of Urban Scenes (2017, background): “Traditional works in semantic segmentation [37, 42, 48] are typically based on manually designed image features.”
Harmony potentials for joint classification and segmentation (2010, methodology): “The incorporation of global information through the harmony potential is novel with respect to existing techniques exploiting image-level priors such as [19].”