Interactive graph cuts for optimal boundary & region segmentation of objects in N-D images

By Yuri Boykov, M. Jolly. 2001

4,598 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

SimpleClick: Interactive Image Segmentation with Simple Vision Transformers (2022, background): “Early works [6, 16, 18, 40] tackle this problem using graphs de-ﬁned over image pixels.”
3D Instance Segmentation of MVS Buildings (2021, methodology): “With the aforementioned MRF setting, we treat Equation 8 as a classical max-flow/min-cut optimization and solve it using the graph cut algorithm [54].”
Reviving Iterative Training with Mask Guidance for Interactive Segmentation (2021, methodology): “Scribbles are used in many early works [5, 15, 16, 17, 26].”
Getting to 99% Accuracy in Interactive Segmentation (2020, methodology): “[6] VGG16 [23] FCN [17] 1:8 graph-cuts [1] ITIS Mahadevan et al.”
Iteratively Trained Interactive Segmentation (2018, methodology): “Graph cut [3] >20 >20 Geodesic matting [1] >20 >20 Random walker [13] 16.”
Deep Extreme Cut: From Extreme Points to Object Segmentation (2017, background): “One of the most common ways to perform weakly supervised segmentation is drawing a bounding box around the object of interest [4, 38, 32, 18].”
Extreme Clicking for Efficient Object Annotation (2017, methodology): “ potential V. The summation over (p;q) in (1) is deﬁned on an eight-connected pixel grid. Usually, this penalty depends on the RGB difference between pixels, being smaller in regions of high contrast [8,6,27,42,51,64]. In this paper, we instead use the sum of the edge responses of the two pixels giv”
Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps (2013, methodology): “quantile of the saliency distribution in the image; the background model was estimated from the pixels with the saliency smaller than the 30% quantile (Fig.3, right-middle). The GraphCut segmentation [3] was then performed using the publicly available implementation2. Once the image pixel labelling ”
ScribbleSup: Scribble-Supervised Convolutional Networks for Semantic Segmentation (2016, methodology): “Formulation in this form is widely used for interactive image segmentation [4, 26, 17, 11, 16, 19].”
An experimental comparison of min-cut/max- flow algorithms for energy minimization in vision (2001, methodology): “We also onsider image restoration with \linear" intera tions energy: E(I) =X p2P jjIp Io p jj+ X (p;q)2N A(p;q) jIp Iq j (3)”
Instance-Level Segmentation for Autonomous Driving with Deep Densely Connected MRFs (2015, methodology): “[3] relies on scribbles as seeds to model appearance of foreground/background, and uses graph-cuts by combining appearance cues and a smoothness term [4].”
A comparative evaluation of interactive segmentation algorithms (2010, neutral): “Different segmentation algorithms are often created with different application domains in mind , and thus suited to different tasks .”
Interactive graph cut based segmentation with shape priors (2005, methodology): “We refer the interested to [2] for a formal proof of this fact.”
MultiSeg: Semantically Meaningful, Scale-Diverse Segmentations From Minimal User Input (2019, methodology): “Following [23, 36], we apply graph cut [4] to the network prediction to obtain the final segmentation mask.”
Segmentation by transduction (2008, methodology): “Besides, the discrete version of our algorithm for s = 2 (equivalently λ = 0 in (3)) is comparable to the regularization used in [4, 5, 6].”
Scene parsing with Multiscale Feature Learning, Purity Trees, and Optimal Covers (2012, background): “In the next two sections we describe these two steps.”