Normalized Cut Meets MRF

Tang, Meng; Marin, Dmitrii; Ayed, Ismail Ben; Boykov, Yuri

doi:10.1007/978-3-319-46475-6_46

Cited by 19 publications

(44 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The first regularization term leverages the pattern of neighboring nodes that have high possibility of sharing same label in a graph. This regularization was realized by using the Markov Random Field (MRF) (Tang et al, 2016;Kohli et al, 2009). The combination of Ncut and MRF so far has been only used to process static images (Tang et al, 2016), not to parcellate the brain through processing BOLD signal time courses in rs-fMRI data.…”

Section: Discussionmentioning

confidence: 99%

“…Distance as a strict constraint may not be suitable for brain network parcellation since some networks (e.g., the default mode network) consist of several disjoint regions. Markov Random Field (MRF) defined over a graph 8 (Tang et al, 2016;Kohli et al, 2009) and balanced by a tuning parameter r was used to cluster spatially contiguous voxels into the network while r was adjusted so that disjoint regions could be clustered into the same network. Weighted MRF was one regularization adopted in the RNcut.…”

Section: Regularized-ncut (Rncut)mentioning

confidence: 99%

“…where Ak denotes the k different sub-graphs or networks; and EV and EP are the two regularization terms further formulated below. EV was defined by using MRF regularization (Tang et al, 2016):…”

Section: Regularized-ncut (Rncut)mentioning

confidence: 99%

See 2 more Smart Citations

Regularized-Ncut: Robust and homogeneous functional parcellation of neonate and adult brain networks

Peng

Ouyang

Wang

et al. 2019

Preprint

View full text Add to dashboard Cite

AbstractBrain network parcellation based on resting-state functional MRI (rs-fMRI) is affected by noise, resulting in spurious small patches and decreased functional homogeneity within each network. Robust and homogeneous parcellation of neonate brain is more difficult with neonate rs-fMRI associated with higher level of noise and no functional atlas as spatial constraints. To meet these challenges, we developed a novel data-driven Regularized Normalized-cut (RNcut) method. RNcut is formulated by adding two regularization terms, a smoothing term using Markov random fields and a small-patch removal term, to conventional normalized-cut (Ncut) method. The RNcut and competing methods were tested with simulated datasets with known ground truth and then applied to both adult and neonate rs-fMRI datasets. Based on the parcellated networks generated by RNcut, intra-network connectivity was quantified. The test results from simulated datasets demonstrated that the RNcut method is more robust (p<0.01) to noise and can delineate parcellated functional networks with significantly better (p<0.01) spatial contiguity and significantly higher (p<0.01) functional homogeneity than competing methods. Application of RNcut to neonate and adult rs-fMRI dataset revealed distinctive functional brain organization of neonate brains from that of adult brains. Collectively, we developed a novel data-driven RNcut method by integrating conventional Ncut with two regularization terms, generating robust and homogeneous functional parcellation without imposing spatial constraints. A broad range of brain network applications and analyses, especially exploratory investigations of parcellating neonate and infant brain with noisy dataset, can potentially benefit from this RNcut method.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Regularized-ncut (Rncut)mentioning

confidence: 99%

See 1 more Smart Citation

Regularized-Ncut: Robust and homogeneous functional parcellation of neonate and adult brain networks

Peng

Ouyang

Wang

et al. 2019

Preprint

View full text Add to dashboard Cite

show abstract

“…On the other hand, we also present the performance gap between weak and full mask training to provide a more fair comparison in Table 2. In Table 2, the performance results of full mask training (64.1 %), GrabCut [22], NormalizedCut [23], and KernelCut [24] are acquired from [23].…”

Section: Methodsmentioning

confidence: 99%

“…mIoU Gap between full and weak supervision With Full Masks [23] 64.1 -GrabCut [22] 55.5 8.6 NormalizedCut [23] 58.7 5.4 KernelCut [24] 59 training (61.6%) by 2.4% and 1.38% at selection of σ FH = 0.8 and σ FHBest , respectively. Two example images from the generated set, i.e.…”

Section: Methodsmentioning

confidence: 99%

Weakly Supervised Semantic Segmentation Using Constrained Dominant Sets

Aslan

Pelillo

2019

Lecture Notes in Computer Science

View full text Add to dashboard Cite

The availability of large-scale data sets is an essential prerequisite for deep learning based semantic segmentation schemes. Since obtaining pixel-level labels is extremely expensive, supervising deep semantic segmentation networks using low-cost weak annotations has been an attractive research problem in recent years. In this work, we explore the potential of Constrained Dominant Sets (CDS ) for generating multilabeled full mask predictions to train a fully convolutional network (FCN) for semantic segmentation. Our experimental results show that using CDS's yields higher-quality mask predictions compared to methods that have been adopted in the literature for the same purpose.

show abstract

Rail‐Lane Extraction Based on Cutting‐Edge Deep Neural Networks and Graph‐Based Segmentation

Yasue,

Futagami

2024

IEEJ Transactions Elec Engng

View full text Add to dashboard Cite

This study aims to enhance the accuracy of rail‐lane extraction, which determines the rail lane and its background at a pixel‐wise level from an ego‐perspective image captured by the front camera of a train. For this purpose, cutting‐edge deep neural network (DNN)‐based semantic segmentation models, MANet and UNet++, were applied. In addition, a method that effectively enlarges the rail lane extracted by DNN‐based segmentation was proposed based on thoroughly analyzing the issues of DNN‐based segmentation. Our experiment, which employed 601 ego‐perspective images, demonstrated that the proposed method using the cutting‐edge DNN‐based segmentation method increased the balanced accuracy by 6.66% compared with the conventionally employed ERFNet in the literature. Further discussions and useful analyses are presented herein. © 2024 Institute of Electrical Engineer of Japan and Wiley Periodicals LLC.

show abstract

Normalized Cut Meets MRF

Cited by 19 publications

References 44 publications

Regularized-Ncut: Robust and homogeneous functional parcellation of neonate and adult brain networks

Regularized-Ncut: Robust and homogeneous functional parcellation of neonate and adult brain networks

Weakly Supervised Semantic Segmentation Using Constrained Dominant Sets

Rail‐Lane Extraction Based on Cutting‐Edge Deep Neural Networks and Graph‐Based Segmentation

Contact Info

Product

Resources

About