Interactive Image Segmentation via Pairwise Likelihood Learning

Authors: Tao Wang, Quansen Sun, Qi Ge, Zexuan Ji, Qiang Chen, Guiyu Xia

IJCAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on challenging data sets demonstrate that the proposed method can obtain better performance than state-of-the-art methods.The proposed method was experimentally verified by comparing it with state-of-the-art approaches: Grab Cut [Rother et al., 2004], RW [Grady, 2006], LC [Casaca et al., 2014], SMRW [Dong et al., 2016] and NHO [Kim et al., 2010] on the Berkeley segmentation data set1 which contains 500 images with size 481 321 (or 321 481 ) and Microsoft Grab Cut database2 which contains 50 images with public seeds information.
Researcher Affiliation Academia 1School of Computer Science and Engineering, Nanjing University of Science and Technology, China 2School of Telecommunications and Information Engineering, Nanjing University of Posts and Telecommunications, China
Pseudocode No Not found. The method is described mathematically and textually, but no structured pseudocode or algorithm blocks are provided.
Open Source Code No Not found. No explicit statement or link for open-source code is provided.
Open Datasets Yes The proposed method was experimentally verified by comparing it with state-of-the-art approaches: Grab Cut [Rother et al., 2004], RW [Grady, 2006], LC [Casaca et al., 2014], SMRW [Dong et al., 2016] and NHO [Kim et al., 2010] on the Berkeley segmentation data set1 which contains 500 images with size 481 321 (or 321 481 ) and Microsoft Grab Cut database2 which contains 50 images with public seeds information. 1http://www.eecs.berkeley.edu/Research/Projects/CS/vision/gr ouping/segbench/S. 2http://research.microsoft.com/enus/um/cambridge/projects/visi onimagevideoediting/segmentation/grabcut.htm
Dataset Splits No Not found. The paper mentions datasets but does not specify explicit training, validation, or test splits for reproducibility.
Hardware Specification Yes Table 3 lists the average running times of Grab Cut, RW, LC, SMRW, NHO and the proposed method on all 20 test images with size 321 481 in the Microsoft Grab Cut database on an Intel Xeon CPU running at 2.0 GHz in MATLAB.
Software Dependencies No Not found. The paper mentions
Experiment Setup Yes Parameters involved in the proposed scheme are set as follows: the constant is fixed as 60, the number of clusters F K and B K are both set to 3, the controlling parameter is set to 0.1, and the 4-neighborhood relationship of pixels is considered.