KSKyohoon Sim

Kyohoon Sim

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Kyohoon Sim has a total of 11 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Blind Stereoscopic Image Quality Evaluator Based on Binocular Semantic and Quality Channels

    by Kyohoon Sim, Jiachen Yang, Wen Lu, Xinbo Gao

    Human beings always evaluate the perceptual quality of an image coupled with identifying the semantic content of images. This paper addresses the correlation issue between stereoscopic image quality assessment (SIQA) and semantic recognition. In contrast to the previous SIQA methods that relied on binocular quality-aware features of a stereoscopic image, our approach also extracts binocular semantic features using a pre-trained deep convolutional neural network (DCNN) on a large dataset like ImageNet dataset, as well as the manually designed binocular quality-aware features. It can solve the problem of limited SIQA dataset size and facilitate better prediction on the quality. Experimental results demonstrate that the binocular semantic features are a good predictor for the stereoscopic image quality. The proposed method outperforms the state-of-the-art SIQA methods on four benchmark SIQA datasets. Significantly, all Spearman rank-order correlation coefficients (SROCCs) between the predicted scores and the subjective scores on the four datasets exceed 0.95. The MATLAB source code of the proposed method is available at https://github.com/kyohoonsim/Blind-Stereoscopic-Image-Quality-Evaluator-based-on-Binocular-Semantic-and-Quality-Channels https://github.com/kyohoonsim .


  • MaD-DLS: Mean and Deviation of Deep and Local Similarity for Image Quality Assessment

    by Kyohoon Sim, Jiachen Yang, Wen Lu, Xinbo Gao

    When human visual system (HVS) looks at a scene, it extracts various features from the image about the scene to understand it. The extracted features are compared with the stored memory on the analogous scene to judge their similarity [1]. By analyzing to the similarity, HVS understands the scene presented on eyes. Based on the neurobiological basis, we propose a 2D full reference (FR) image quality assessment (IQA) method, named mean and deviation of deep and local similarity (MaD-DLS) that compares similarity between many original and distorted deep feature maps from convolutional neural networks (CNNs). MaD-DLS uses a deep learning algorithm, but since it uses the convolutional layers of a pre-trained model, it is free from training. For pooling of local quality scores within a deep similarity map, we employ two important descriptive statistics, (weighted) mean and standard deviation and name it mean and deviation (MaD) pooling. The two statistics each have the physical meaning: the weighted mean reflects effect of visual saliency on quality, whereas the standard deviation reflects effect of distortion distribution within the image on it. Experimental results show that MaD-DLS is superior or competitive to the existing methods and the MaD pooling is effective. The MATLAB source code of MaD-DLS will be available online soon (after the manuscript is accepted).


  • No-reference image quality assessment focusing on human facial region

    by Jiachen Yang, Zhihui Huang, Kyohoon Sim, Wen Lu, Kai Liu, Hehan Liu


  • 3-D Visual Discomfort Assessment Considering Optical and Neural Attention Models

    by Jiachen Yang, Vanhung Nguyen, Kyohoon Sim, Yang Zhao, Wen Lu

    In this paper, we find factors affecting visual comfort along the visual pathways. The statistics data used for the 3D visual discomfort predictor (3D-VDP) model extracted from the traditional optical statistical elements derived, retinal disparity and neurophysiological elements obtained from V1 and MT areas. These physiological statistics are used to create a new feature map for 3D-VDP models, known as MT importance map. The thirteen typical MT neurons are divided into four kinds of cells: tuned-excitatory cells, tuned-inhibitory cells, far cells and near cells based on their characteristics. Disparity map is also transformed by changing the viewing distance in the formula to calculate the horizontal angular disparity factor. We propose that variation of viewing distance will make the MT cells more intense reaction. As a result, three types of angular disparity maps are obtained: tuned-near, tuned-far and original disparity maps. Preliminary results on two public available databases exhibit the performance of the proposed approach outperforms previous S3D VDP models.


  • Predicting Stereoscopic Image Quality via Stacked Auto-Encoders Based on Stereopsis Formation

    by Jiachen Yang, Kyohoon Sim, Wen Lu, Bin Jiang

    Most of previous 2D and 3D image quality evaluators were based on shallow architectures. Their shallow architectures cannot model phenomenon occurring in human visual system (HVS) sufficiently. Disparities between left and right views have been importantly used for 3D image quality assessment (IQA), but single disparity, depth or cyclopean maps made from the disparities cannot thoroughly reflect depth sense. In this paper, we propose a blind stereoscopic image quality evaluator using Stacked Auto-Encoders (SAE). The proposed method is based on two theories on initial stages of stereopsis. One is cyclopean channel theory, the other is binocular summation/difference channels theory. Especially a cyclopean image which models the former theory is computed to consider binocular suppression, whereas summation and difference images which model the latter one are utilized to treat the depth sense. We train three SAEs for quality-aware features from the three images in an unsupervised manner. Through the SAEs, the features are transformed into more meaningful features, and they are used to train two regressors. The regressors can be used to get a final predicted score. Experimental results conducted on popular 3D IQA databases prove that the proposed algorithm outperforms state-of-the-art 3D IQA methods.