Article
Computer Science, Artificial Intelligence
Ling Li, Chunyi Chen, Jun Peng, Ripei Zhang
Summary: This paper proposes a model called Human Visual Perception and Deep Learning Image Difference Metric (HPDL-IDM), which combines the Human Visual System (HVS) model and deep learning, to evaluate the visibility difference between distorted images and reference images at a pixel-wise level. Experimental results show that HPDL-IDM exhibits significantly improved generalization capacity and accuracy compared to other models.
IET IMAGE PROCESSING
(2023)
Article
Computer Science, Artificial Intelligence
Yannick Hold-Geoffroy, Dominique Piche-Meunier, Kalyan Sunkavalli, Jean-Charles Bazin, Francois Rameau, Jean-Francois Lalonde
Summary: Image editing and compositing are widely used in entertainment, and require geometric camera calibration for beautiful composites. This paper proposes using a deep convolutional neural network to directly infer camera calibration parameters from a single image. A human perception study is conducted to investigate sensitivity to calibration inaccuracies, and a new perceptual measure is developed. The deep calibration network outperforms previous methods and is applied to various applications.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Xuelin Shen, Zhangkai Ni, Wenhan Yang, Xinfeng Zhang, Shiqi Wang, Sam Kwong
Summary: This paper proposes an effective approach to infer the just noticeable distortion (JND) profile based on patch-level structural visibility learning, and trains a deep learning-based structural degradation estimation model. The method demonstrates superiority over existing techniques, and establishes a corresponding JND dataset for the learning process.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2021)
Article
Computer Science, Software Engineering
Cara Tursun, Piotr Didyk
Summary: Modeling the temporal aspect of visual perception in the periphery is crucial for optimizing and evaluating content generation techniques. This article presents psychophysical experiments measuring human observers' sensitivity to spatio-temporal stimuli across a wide field of view and builds a perceptual model using the collected data. The model enables injecting new content into the periphery without distracting the viewer and optimizes foveated rendering methods to limit the visibility of temporal aliasing.
ACM TRANSACTIONS ON GRAPHICS
(2023)
Article
Multidisciplinary Sciences
Lauren V. Hadley, Jamie A. Ward
Summary: The study found that head movement synchrony increases in high noise levels as opposed to low noise levels during conversation, and that there is greater movement coherence in multi-talker babble compared to speech-shaped noise in triads. Differences in synchrony predominantly fall in the 0.2-0.5 Hz frequency bands, with additional differences at higher frequencies in triads only (>5 Hz), possibly related to backchannel cues synchrony. This research replicates prior findings of enhanced reliance on behavioral synchrony as task difficulty increases, using various difficulty manipulations and group sizes.
Article
Automation & Control Systems
Miaohui Wang, Yijing Huang, Jian Xiong, Wuyuan Xie
Summary: This article introduces a new visibility perception-guided blind quality indicator for low-light images in-the-wild, considering visibility perception, luminosity cognition, and color sensation. Experimental results show that the proposed indicator outperforms nine representative methods.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Article
Engineering, Electrical & Electronic
Shuai Liu, Yingcong Huang, Huoxiang Yang, Yongsheng Liang, Wei Liu
Summary: In recent years, image compression methods based on deep learning have attracted extensive attention. However, traditional methods that minimize mean squared error are limited in capturing perceptual differences between images, resulting in poor visual quality of the reconstructed images. To address this issue, we propose a rate-distortion loss based on perception metric to enhance the visual perception quality of the reconstructed images.
SIGNAL IMAGE AND VIDEO PROCESSING
(2022)
Article
Computer Science, Information Systems
Pengfei Guo, Lang He, Shuangyin Liu, Delu Zeng, Hantao Liu
Summary: This paper investigates the performance of five popular enhancement algorithms for underwater images and analyzes their impact on perceptual quality. It also evaluates the visual quality objectively, aiming to develop objective metrics for automatic assessment of underwater image enhancement quality.
IEEE TRANSACTIONS ON MULTIMEDIA
(2022)
Article
Neurosciences
Carmen Krewer, Lea John, Jeannine Bergmann, Stanislav Bardins, Klaus Jahn
Summary: The study compared two methods of measuring subjective postural vertical (SPV) and found that they produced similar error values but significantly different range values. Both methods can be used interchangeably within the same study.
NEUROSCIENCE LETTERS
(2021)
Article
Computer Science, Software Engineering
Yana Nehme, Florent Dupont, Jean-Philippe Farrugia, Patrick Le Callet, Guillaume Lavoue
Summary: This study introduces a large dataset of 480 animated meshes with diffuse color information and proposes the first metric for quality assessment of 3D meshes with diffuse colors, achieving good results and stability. Additionally, the research investigates how knowledge of the viewpoint can improve the results of objective metrics.
IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS
(2021)
Article
Engineering, Electrical & Electronic
Liqiang Guo, Lian Liu
Summary: This letter proposes a perceptual-based robust focus measure based on the local difference of Gaussian (DoG), which can quantify the sharpness information of images under noise conditions and outperforms other competing methods in terms of robustness.
IEEE SIGNAL PROCESSING LETTERS
(2022)
Article
Engineering, Electrical & Electronic
Hanhe Lin, Guangan Chen, Mohsen Jenadeleh, Vlad Hosu, Ulf-Dietrich Reips, Raouf Hamzaoui, Dietmar Saupe
Summary: PJND is the smallest distortion level that a subject can perceive when an image is compressed. Accurate and diverse results require a large number of subjects and images. To address the limitations of laboratory experiments, researchers have developed a framework for PJND assessments via crowdsourcing.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2022)
Article
Medicine, General & Internal
Carlos Alberto Leite Filho, Monica de Oliveira Viana, Fatima Cristina Alves Branco-Barreiro, Silvana Maria Monte Coelho Frota
Summary: The aim of this study was to investigate the effect of age on the performance of children aged 7-12 years in the MLD test and establish reference values for this age group. The findings indicate that age does not significantly affect test performance.
JOURNAL OF CLINICAL MEDICINE
(2022)
Article
Psychology, Experimental
Jason Samaha, Joshua J. LaRocque, Bradley R. Postle
Summary: The amplitude of spontaneous oscillatory alpha waves is negatively related to subjective visibility judgments in perception of high-level visual stimuli, but does not predict trial-by-trial accuracy.
CONSCIOUSNESS AND COGNITION
(2022)
Article
Computer Science, Information Systems
Hui Men, Hanhe Lin, Mohsen Jenadeleh, Dietmar Saupe
Summary: The study utilizes boosting techniques and triplet comparisons to evaluate the quality of distorted images. Boosting increases sensitivity, leading to more accurate subjective assessments while reducing the number of subjective ratings. Experimental results demonstrate the effectiveness of this approach across various types of distortion.
Article
Computer Science, Software Engineering
L. Vasa, P. Vanecek, M. Prantl, V. Skorkovska, P. Martinek, I. Kolingerova
COMPUTER GRAPHICS FORUM
(2016)
Article
Computer Science, Software Engineering
Guillaume Lavoue, Mohamed Chaker Larabi, Libor Vasa
IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS
(2016)
Article
Computer Science, Software Engineering
M. Prantl, L. Vasa
Article
Computer Science, Software Engineering
L. Hruda, J. Dvorak, L. Vasa
COMPUTER GRAPHICS FORUM
(2019)
Article
Biochemical Research Methods
Jan Dvorak, Martin Manak, Libor Vasa
JOURNAL OF MOLECULAR GRAPHICS & MODELLING
(2020)
Article
Computer Science, Software Engineering
Lukas Hruda, Ivana Kolingerova, Libor Vasa
Summary: Reflectional symmetry is a significant feature exhibited by many real-world objects. This paper introduces a novel differentiable symmetry measure and a method for symmetry plane detection in 3D objects, which performs well and is robust. The proposed method was tested with good results and outperformed other state-of-the-art methods.
Article
Computer Science, Software Engineering
Jan Dvorak, Zuzana Kacerekova, Petr Vanecek, Lukas Hruda, Libor Vasa
Summary: This paper discusses an improved methodology for analyzing and computing time-consistent correspondence by tracking the volume within the surface. By modifying the energy formulation, incorporating a notion of volume element affinity, and using an improved optimization strategy, better results are achieved. Additionally, two metrics are introduced to quantify the quality of the tracking results.
COMPUTERS & GRAPHICS-UK
(2022)
Article
Computer Science, Software Engineering
Martin Manak, Alexey Anikeenko, Libor Vasa, Ivana Kolingerova
Summary: Voronoi diagrams and their dual tetrahedral structures are commonly used in protein model analysis. However, compressing the dual tetrahedral structure of aw-VDs poses unique challenges. This study proposes a method to compactly store the dual tetrahedral structure and demonstrates its effectiveness on protein data.
Article
Computer Science, Software Engineering
Jan Dvorak, Zuzana Kacerekova, Petr Vanccek, Libor Vasa
Summary: In this paper, an improved algorithm is proposed for encoding the connectivity of a triangle mesh given a known geometry. By replacing the fixed traversal with a dynamic selection of the best prediction using a priority queue and quality metric, a significant reduction in the required data rate is achieved compared to existing techniques.
COMPUTER GRAPHICS FORUM
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Martin Cervenka, Ondrej Havlicek, Josef Kohout, Libor Vasa
Summary: The main challenges of collision detection and handling in muscle modelling are demonstrated in this paper. A collision handling technique is tested to address the issue of muscle penetrating the bone in certain circumstances. Alternative techniques to the described PBD technique are also presented.
PROCEEDINGS OF THE 18TH INTERNATIONAL JOINT CONFERENCE ON COMPUTER VISION, IMAGING AND COMPUTER GRAPHICS THEORY AND APPLICATIONS, VISIGRAPP 2023
(2023)
Article
Computer Science, Information Systems
Gerasimos Arvanitis, Evangelia I. Zacharaki, Libor Vasa, Konstantinos Moustakas
Summary: This study presents a method for registering and identifying partially-scanned and noisy 3D objects, and matching them with high-quality models. The method performs well in dealing with point cloud scenes with significant missing parts.
IEEE TRANSACTIONS ON MULTIMEDIA
(2022)
Proceedings Paper
Computer Science, Cybernetics
Eliska Mourycova, Libor Vasa
Summary: This paper discusses lossy geometry compression of manifold triangle meshes using the EdgeBreaker algorithm and a known reference shape. The predictions of vertices positions are made outside the reference shape and then orthogonally projected onto its surface. The corrections are encoded using two integer numbers. Compared to a state-of-the-art static mesh compression algorithm, this approach with a reference mesh requires a smaller bitrate for comparable error.
GRAPP: PROCEEDINGS OF THE 17TH INTERNATIONAL JOINT CONFERENCE ON COMPUTER VISION, IMAGING AND COMPUTER GRAPHICS THEORY AND APPLICATIONS - VOL 1: GRAPP
(2022)
Article
Computer Science, Software Engineering
L. Vasa, J. Dvorak
COMPUTER GRAPHICS FORUM
(2018)
Proceedings Paper
Computer Science, Artificial Intelligence
Martin Prantl, Libor Vasa, Ivana Kolingerova
COMPUTER VISION, IMAGING AND COMPUTER GRAPHICS THEORY AND APPLICATIONS, VISIGRAPP 2016
(2017)
Proceedings Paper
Geography
Karel Janecka, Libor Vasa
RISE OF BIG SPATIAL DATA
(2017)