Article
Computer Science, Artificial Intelligence
Rosaura G. VidalMata, Sreya Banerjee, Brandon RichardWebster, Michael Albright, Pedro Davalos, Scott McCloskey, Ben Miller, Asong Tambo, Sushobhan Ghosh, Sudarshan Nagesh, Ye Yuan, Yueyu Hu, Junru Wu, Wenhan Yang, Xiaoshuai Zhang, Jiaying Liu, Zhangyang Wang, Hwann-Tzong Chen, Tzu-Wei Huang, Wen-Chi Chin, Yi-Chun Li, Mahmoud Lababidi, Charles Otto, Walter J. Scheirer
Summary: The current state-of-the-art for image restoration and enhancement in degraded images acquired under less than ideal circumstances faces the challenge of translating capabilities to useful visual recognition. Development of algorithms designed to improve visual appearance and recognition is necessary, with large-scale datasets and evaluation metrics playing a key role in driving innovation in this area.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2021)
Article
Computer Science, Software Engineering
Zeqiang Lai, Kaixuan Wei, Ying Fu, Philipp Hartel, Felix Heide
Summary: This article introduces del-Prox, a domain-specific modeling language and compiler for large-scale optimization problems using differentiable proximal algorithms. del-Prox's core feature is full differentiability, supporting hybrid model- and learning-based solvers that integrate proximal optimization with neural network pipelines. With just a few lines of code, del-Prox can generate performant solvers for various image optimization problems, including end-to-end computational optics, image deraining, and compressive magnetic resonance imaging. It can also be used in completely different domains such as energy system planning, outperforming state-of-the-art CVXPY and commercial Gurobi solvers.
ACM TRANSACTIONS ON GRAPHICS
(2023)
Article
Mathematics
Yenny Villuendas-Rey, Jose L. Velazquez-Rodriguez, Mariana Dayanara Alanis-Tamez, Marco-Antonio Moreno-Ibarra, Cornelio Yanez-Marquez
Summary: When facing challenges in science, engineering or technology, it is essential to seek the best solution through optimization. The Mexican Axolotl Optimization algorithm, inspired by nature, achieved very good optimization results in most experiments.
Article
Computer Science, Information Systems
Yang Yang, Dan Wu, Lanling Zeng, Zhuoran Li
Summary: The proposed filter is a novel edge-preserving filter based on deep unsupervised learning under the weighted least square framework. It is capable of efficiently smoothing images and suppressing artifacts.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Software Engineering
Yang Yang, Dan Wu, Ling Tang, Lanling Zeng, Zhigeng Pan
Summary: In this paper, a weighted and truncated L-1-regularized optimization model is proposed for image smoothing. A deep unsupervised learning-based filter is also proposed based on the defined loss function. The experimental results show that the proposed filter outperforms state-of-the-art filters in terms of image quality and is highly efficient.
Article
Multidisciplinary Sciences
Taewon Cho, Hodjat Pendar, Julianne Chung
Summary: The paper investigates the inverse problem of recovering gas exchange signals of animals placed in a flow-through respirometry chamber from measured gas concentrations, which is computationally challenging for large-scale experiments with uncertainties. Various computational tools and nonlinear optimization methods are described for reconstruction and uncertainty quantification, addressing both known and unknown impulse response functions. Numerical experiments demonstrate the benefits and potential impacts of these methods in respirometry.
Article
Computer Science, Software Engineering
Steven Diamond, Vincent Sitzmann, Frank Julca-Aguilar, Stephen Boyd, Gordon Wetzstein, Felix Heide
Summary: The traditional imaging process involves a series of steps, while advanced processing includes feature extraction, classification, tracking, and fusion. We propose an end-to-end differentiable architecture that performs demosaicking, denoising, deblurring, tone-mapping, and classification, improving perception in low light and challenging conditions.
ACM TRANSACTIONS ON GRAPHICS
(2021)
Article
Geochemistry & Geophysics
Ronghuo Dai, Cheng Yin, Da Peng
Summary: Sparsity-regularized linear inverse problems are widely used in various fields, and the iterative hard thresholding algorithm and iterative soft thresholding algorithm are frequently used methods to solve them. This letter proposes a hybrid method that incorporates Davidon-Fletcher-Powell formulations into iterative thresholding-like algorithms to improve convergence rate. Numerical examples and real-life application demonstrate its effectiveness.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2023)
Article
Computer Science, Artificial Intelligence
Seng Cheong Loke, Bruce A. MacDonald, Matthew Parsons, Burkhard Claus Wunsche
Summary: In this article, a new routine for superpixel image segmentation (F-DBSCAN) based on the DBSCAN algorithm is described, which is six times faster than previous methods. The program achieves faster speeds through efficient parallelization of cluster search processes and improved performance through calculations in large consolidated memory buffers. Tested on the Berkeley Segmentation Dataset, the average processing speed is 175 frames/s with a Boundary Recall of 0.797 and an Achievable Segmentation Accuracy of 0.944.
JOURNAL OF REAL-TIME IMAGE PROCESSING
(2021)
Article
Engineering, Electrical & Electronic
Shota Takahashi, Mirai Tanaka, Shiro Ikeda
Summary: This paper introduces a technique for blind deconvolution, which recovers an original signal by minimizing a quartic objective function. The Bregman-based proximal methods are employed, and their effectiveness and superior performance over existing algorithms are demonstrated through numerical experiments.
Article
Computer Science, Software Engineering
Praneeth Chakravarthula, Ethan Tseng, Henry Fuchs, Felix Heide
Summary: Holography is a promising avenue for high-quality displays, but existing limitations such as wavefront modeling and unit element encoding have hindered its full potential. The newly proposed hogel-free holography overcomes these limitations and supports true 3D holographic effects.
ACM TRANSACTIONS ON GRAPHICS
(2022)
Article
Multidisciplinary Sciences
Karsten Wuellems, Annika Zurowietz, Martin Zurowietz, Roland Schneider, Hanna Bednarz, Karsten Niehaus, Tim W. Nattkemper
Summary: Mass Spectrometry Imaging (MSI) is a technique for spatial analysis of molecular co-location in biological samples that is expanding into new domains such as clinical pathology. In order to improve the value of MSI data, new tools QUIMBI and ProViM have been developed to provide a intuitive and convenient visual analysis method.
SCIENTIFIC REPORTS
(2021)
Article
Optics
Xianglei Liu, Joao Monteiro, Isabela Albuquerque, Yingming Lai, Cheng Jiang, Shian Zhang, Tiago H. Falk, Jinyang Liang
Summary: The paper introduces a snapshot-to-video autoencoder (S2V-AE) for machine-learning assisted real-time processing in compressed ultrahigh-speed imaging, achieving efficient reconstruction time and a large sequence depth.
PHOTONICS RESEARCH
(2021)
Article
Optics
Zhihong Zhang, Chao Deng, Yang Liu, Xin Yuan, Jinli Suo, Qionghai Dai
Summary: This paper introduces a novel hybrid coded aperture snapshot compressive imaging (HCA-SCI) system, which achieves high-resolution high-speed video capture with a throughput of 4.6 x 10^9 voxels per second. Simulations and real-data experiments confirm the feasibility and performance of the proposed system and algorithm.
PHOTONICS RESEARCH
(2021)
Article
Computer Science, Software Engineering
Praneeth Chakravarthula, Seung-Hwan Baek, Florian Schiffers, Ethan Tseng, Grace Kuo, Andrew Maimone, Nathan Matsuda, Oliver Cossairt, Douglas Lanman, Felix Heide
Summary: Holographic displays offer enhanced display capabilities in augmented reality applications, but the sampling of the holographic field by the eye pupil poses a challenge. This study introduces a pupil-aware holography approach that maximizes perceptual image quality regardless of the eye pupil's size, location, and orientation. Experimental results show that the proposed method eliminates severe artifacts and outperforms existing approaches.
ACM TRANSACTIONS ON GRAPHICS
(2022)
Article
Computer Science, Software Engineering
M. Kluge, T. Weyrich, A. Kolb
COMPUTER GRAPHICS FORUM
(2020)
Article
Computer Science, Software Engineering
R. Winchenbach, A. Kolb
COMPUTER GRAPHICS FORUM
(2020)
Article
Computer Science, Software Engineering
Rene Winchenbach, Rustam Akhunov, Andreas Kolb
ACM TRANSACTIONS ON GRAPHICS
(2020)
Article
Computer Science, Software Engineering
Rene Winchenbach, Andreas Kolb
Summary: The article introduces an improved method for simulating low-viscosity turbulent flows with higher adaptive volume ratios using Smoothed Particle Hydrodynamics. By deriving a discretized objective function and optimizing it online, ideal refinement patterns can be generated without intuitive initial user-input. The research reveals a residual refinement error term, which is smoothed out using novel methods to achieve significantly higher adaptive volume ratios than previous work.
ACM TRANSACTIONS ON GRAPHICS
(2021)
Article
Computer Science, Artificial Intelligence
Paramanand Chandramouli, Kanchana Vaishnavi Gandikota, Andreas Gorlitz, Andreas Kolb, Michael Moeller
Summary: This work presents a generative model for 4D light field patches using variational autoencoders, serving as a prior for light field reconstruction tasks. The proposed method achieves better performance than traditional model-based approaches on both synthetic and real scenes.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2022)
Article
Imaging Science & Photographic Technology
Dmitri Presnov, Andreas Kolb
Summary: We design and perform a user study to evaluate the perception of periodic contour modifications related to their geometry and colour. Based on the study results, we statistically derive a perceptual model, which demonstrates a mainly linear stimulus-to-perception relationship for geometric and colour amplitude and a close-to-quadratic relationship for the respective frequencies, with distinguishable, visually equidistant quantization levels extracted for each contour-related visual variable.
JOURNAL OF IMAGING
(2022)
Article
Health Care Sciences & Services
Dmitri Presnov, Julia Kurz, Judith Willkomm, Johannes Dillmann, Daniel Alt, Robert Zilke, Veit Braun, Cornelius Schubert, Andreas Kolb
Summary: The workflow in modern hospitals requires efficient communication of patient-related medical data to colleagues. This paper introduces a novel concept of anatomically integrated in-place visualization, using a virtual patient's body as a spatial representation of abstract medical data. A prototype on a mobile device for the diagnosis of spinal disc herniation has been evaluated positively by neurosurgeons, who found benefits in the anatomical integration, such as intuitiveness and improved data availability.
HEALTH INFORMATICS JOURNAL
(2023)
Article
Computer Science, Software Engineering
Hendrik Sommerhoff, Andreas Kolb
Summary: In this paper, we propose an A-buffer-based approach that directly renders unprocessed point clouds and filters out outliers in real time, without introducing additional render passes. Our method significantly improves visual quality when rendering noisy point clouds with varying noise levels and large outliers, while only requiring little performance and memory overhead compared to traditional point rendering methods.
Article
Computer Science, Software Engineering
Markus Kluge, Tim Weyrich, Andreas Kolb
Summary: In recent years, there has been an increase in the computing power of mobile camera devices, leading to the development of digitization algorithms that fuse camera observations into a progressively updated scene representation. Previous algorithms either focus on 3D surface representations using depth maps or reconstruct 2D images with respect to a single reference view. Our work combines aspects of both, reconstructing a 2.5-D representation (color and depth) from a fixed viewpoint with variable resolution. This approach provides improved robustness, fidelity, and storage efficiency compared to general 3D reconstructions.
COMPUTERS & GRAPHICS-UK
(2023)
Article
Computer Science, Software Engineering
Dmitri Presnov, Marlena Berels, Andreas Kolb
Summary: This paper introduces the parametric contour-based modification (PACEMOD) approach, which lays the foundations for automated and controllable icon manipulations for the design and visualization of metaphoric glyphs. The PACEMOD parametric representation utilizes diffusion curves with new degrees of freedom in arc-length parameterization to manipulate the geometry and color attributes of the icon contours. Furthermore, the practicality of periodic contour modifications is demonstrated through two visualization examples, including uncertainty visualization of a rain forecast and gradient glyphs for COVID-19 data. In summary, the PACEMOD approach provides a tool for user-centered design of metaphoric glyphs and serves as a generic basis for potential further implementations in specific applications.
Article
Imaging Science & Photographic Technology
Hendrik Sommerhoff, Andreas Kolb
Summary: This paper proposes a generic depth-refinement scheme based on GeoNet, which utilizes the inherent geometric relationship between depth and normal maps to guide a neural network. The approach achieves good results in various depth reconstruction tasks.
JOURNAL OF IMAGING
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Ch Pomrehn, A. Kolb, R. Herpers
Summary: This paper introduces a new methodology for the automated estimation of grayscale representations for hyperspectral images in the context of multimodal vibrational microspectroscopic imagery. It demonstrates that the proposed approach, which derives and fuses gradient information globally in the spectral domain, improves device-based registration of HSI generated by RMS and IRMS.
2021 11TH WORKSHOP ON HYPERSPECTRAL IMAGING AND SIGNAL PROCESSING: EVOLUTION IN REMOTE SENSING (WHISPERS)
(2021)
Article
Social Issues
Cornelius Schubert, Andreas Kolb
Summary: The research focuses on facilitating collaborative activities between computer graphics designers and social scientists in systems design processes, elaborating on a conceptual symmetrical mode of technology design and theory development. Through interdisciplinary engagement in a trading zone, the study highlights the mutual interactions between computer and social scientists, aiming to advance cooperative systems design.
SCIENCE TECHNOLOGY & HUMAN VALUES
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Andreas Goerlitz, Jonas Geiping, Andreas Kolb
2019 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS)
(2019)
Proceedings Paper
Optics
Daniel Stock, Matthias Kahl, Anna K. Wigger, Tak Ming Wong, Andreas Kolb, Peter Haring Bolivar
TERAHERTZ, RF, MILLIMETER, AND SUBMILLIMETER-WAVE TECHNOLOGY AND APPLICATIONS XII
(2019)
