Efficient Graph-Based Image Segmentation

By Pedro F. Felzenszwalb, Dan Huttenlocher. 2004

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

What Makes for Effective Detection Proposals? (2015, methodology): “One approach for overcoming the tension between computational tractability and high detection quality is through the use of “detection proposals” [12]–[15].”
Open-World Instance Segmentation: Exploiting Pseudo Ground Truth From Learned Pairwise Affinity (2022, methodology): “We found that UCM-based methods significantly outperforms other two methods (Table 2): whereas CC and GBH make decision on a merge based on a single pairwise relationship, UCM uses all relations between two segments, and is more robust.”
Fast Edge Detection Using Structured Forests (2014, neutral): “75 15 Felz-Hutt [16] .”
Multi-Object Representation Learning with Iterative Variational Inference (2019, background): “…results on real-world images, however, many approaches (such as “semantic seg-mentation” or object detection) rely on supervised signals (Girshick, 2015; He et al., 2017; Redmon & Farhadi, 2018), while others require hand-engineered features (Shi & Malik, 2000; Felzenszwalb & Huttenlocher, 2004).”
Learning human activities and object affordances from RGB-D videos (2012, methodology): “The other two segmentation methods use the graph-based segmentation proposed by Felzenszwalb and Huttenlocher (2004) adapted to temporally segment the videos.”
Higher Order Conditional Random Fields in Deep Neural Networks (2015, methodology): “The next type of higher order potential we use is based on the idea that superpixels obtained from oversegmentation [1,12] quite often contain pixels from the same visual object.”
Parsing IKEA Objects: Fine Pose Estimation (2013, methodology): “To measure this spill over, we first extract super-pixels, RI , from I using [3] and compute a ratio between an area of a proposed pose, |RP |, and an area of regions that has some overlap with RP .”
DeepEdge: A multi-scale bifurcated deep network for top-down contour detection (2014, background): “A commonly established pipeline in computer vision starts with with low-level contour prediction and then moves up to higher-level object detection.”
Geometric context from a single image (2005, neutral): “[7] to obtain a set of “superpixels”.”
Indoor Semantic Segmentation using depth information (2013, methodology): “We use superpixels [10] to smooth the convnet predictions as a post-processing step, by agregating the classifiers predictions in each superpixel.”
Self-taught object localization with deep networks (2014, methodology): “Specifically, as in [30], we also employ the segmentation method proposed in [11] to generate the initial set ofK rectangular1 regions {b1,b2, .”
Anticipating Human Activities Using Object Affordances for Reactive Robotic Response (2013, methodology): “These include graph based segmentation method proposed by [81] adapted to temporally segment the videos.”