☆ 4.5 Article

Spatially varying image based lighting using HDR-video

COMPUTERS & GRAPHICS-UK (2013)

Journal

COMPUTERS & GRAPHICS-UK

Volume 37, Issue 7, Pages 923-934

Publisher

PERGAMON-ELSEVIER SCIENCE LTD

DOI: 10.1016/j.cag.2013.07.001

Keywords

High dynamic range video; Image based lighting; Scene capture and processing; Photo realistic rendering

Categories

Computer Science, Software Engineering

Funding

Swedish Foundation for Strategic Research (SSF) [IIS11-0081]
Linkoping University Center for Industrial Information Technology (CENIIT)
Swedish Research Council through the Linnaeus Environment CADICS
Swedish Foundation for Strategic Research (SSF) [IIS11-0081] Funding Source: Swedish Foundation for Strategic Research (SSF)

Ask authors/readers for more resources

Protocol

Community support

Reagent

Community support

Abstract

Illumination is one of the key components in the creation of realistic renderings of scenes containing virtual objects. In this paper, we present a set of novel algorithms and data structures for visualization, processing and rendering with real world lighting conditions captured using High Dynamic Range (HDR) video. The presented algorithms enable rapid construction of general and editable representations of the lighting environment, as well as extraction and fitting of sampled reflectance to parametric BRDF models. For efficient representation and rendering of the sampled lighting environment function, we consider an adaptive (2D/4D) data structure for storage of light field data on proxy geometry describing the scene. To demonstrate the usefulness of the algorithms, they are presented in the context of a fully integrated framework for spatially varying image based lighting. We show reconstructions of example scenes and resulting production quality renderings of virtual furniture with spatially varying real world illumination including occlusions. (C) 2013 The Authors. Published by Elsevier Ltd. All rights reserved.

Authors

I am an author on this paper

Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.5

Not enough ratings

Secondary Ratings

Novelty

-

Significance

-

Scientific rigor

-

Rate this paper

Recommended

Article Computer Science, Software Engineering

SkyGAN: Realistic Cloud Imagery for Image-based Lighting

Martin Mirbauer, Tobias Rittig, Tomas Iser, Jaroslav Krivanek, Elena Sikudova

Summary: Achieving photorealism in rendering virtual scenes requires realistic illumination and background. This study presents a hybrid data-driven analytical model that can generate realistic environment maps based on user-defined sun position and cloud coverage ratio.

COMPUTER GRAPHICS FORUM (2023)

Add to Collection

Article Computer Science, Artificial Intelligence

Defocus to focus: Photo-realistic bokeh rendering by fusing defocus and radiance priors

Xianrui Luo, Juewen Peng, Ke Xian, Zijin Wu, Zhiguo Cao

Summary: In this study, a "Defocus to Focus" framework is proposed to learn realistic bokeh rendering. The framework combines defocus priors with the all-in-focus image and implements radiance priors in layered fusion to model depth relations, sustain sharp in-focus regions, and render physically accurate Circle of Confusion (CoC). Extensive experiments on a large-scale bokeh dataset demonstrate that the approach is capable of producing visually pleasing bokeh effects in complex scenes.

INFORMATION FUSION (2023)

Add to Collection

Article Computer Science, Artificial Intelligence

Hand Pose Understanding With Large-Scale Photo-Realistic Rendering Dataset

Xiaoming Deng, Yinda Zhang, Jian Shi, Yuying Zhu, Dachuan Cheng, Dexin Zuo, Zhaopeng Cui, Ping Tan, Liang Chang, Hongan Wang

Summary: Understanding hand poses is crucial in applications like human computer interaction and augmented reality. A lack of high-quality and large-scale dataset has been a barrier for further improvement in hand pose related tasks. This study demonstrates the potential of synthetic datasets in enhancing performance and enabling multi-task learning in hand pose understanding.

IEEE TRANSACTIONS ON IMAGE PROCESSING (2021)

Add to Collection

Article Chemistry, Analytical

Brickognize: Applying Photo-Realistic Image Synthesis for Lego Bricks Recognition with Limited Data

Joel Vidal, Guillem Vallicrosa, Robert Marti, Marc Barnada

Summary: In recent years, supervised deep convolutional neural networks have become the state-of-the-art for image recognition tasks. However, their performance heavily relies on the amount and quality of training data. Acquiring and labeling data is a major challenge, particularly when data is limited. This paper proposes a method using photo-realistic image synthesis and few-shot fine-tuning to overcome limited data in the context of Lego bricks recognition. The results show the effectiveness of synthetically generated data and the successful use of limited data for fine-tuning, achieving high accuracy in both controlled and uncontrolled scenarios.

SENSORS (2023)

Add to Collection

Article Computer Science, Software Engineering

Interactive Photo Editing on Smartphones via Intrinsic Decomposition

Sumit Shekhar, Max Reimann, Maximilian Mayer, Amir Semmo, Sebastian Pasewaldt, Juergen Doellner, Matthias Trapp

Summary: Intrinsic decomposition refers to estimating scene characteristics when given scene views. This research presents a system for intrinsic decomposition of RGB-D images on smartphones, considering both diffuse and specular pixels, and achieving high-quality outputs.

COMPUTER GRAPHICS FORUM (2021)

Add to Collection

Article Automation & Control Systems

Synthetic to Realistic Imbalanced Domain Adaption for Urban Scene Perception

Yining Hua, Dewei Yi

Summary: This article proposes a method to address the imbalance problems in semantic segmentation adaptation by using a progressive hierarchical feature alignment approach, which effectively mitigates the impacts of data and class imbalances.

IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS (2022)

Add to Collection

Article Engineering, Electrical & Electronic

Photo-Realistic Image Super-Resolution via Variational Autoencoders

Zhi-Song Liu, Wan-Chi Siu, Yui-Lam Chan

Summary: A new method for image super-resolution, SR-VAE, is proposed in this study, which utilizes variational autoencoders to learn the conditional distribution of high-resolution images induced by low-resolution images. This method achieves good perceptual quality closer to the natural image manifold while maintaining low distortion.

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY (2021)

Add to Collection

Article Computer Science, Software Engineering

Neural Photo-Finishing

Ethan Tseng, Yuxuan Zhang, Lars Jebe, Xuaner Zhang, Zhihao Xia, Yifei Fan, Felix Heide, Jiawen Chen

Summary: This work presents an approach to make complex photo-finishing pipelines differentiable, allowing legacy algorithms to be trained with first-order optimization methods. By combining tailored network proxy models of individual processing steps, a non-differentiable reference image finishing pipeline can be more faithfully modeled and applied in various applications.

ACM TRANSACTIONS ON GRAPHICS (2022)

Add to Collection

Article Computer Science, Software Engineering

Dynamic scene novel view synthesis via deferred spatio-temporal consistency

Beatrix-Emoke Fulop-Balogh, Eleanor Tursman, James Tompkin, Julie Digne, Nicolas Bonneel

Summary: Structure from motion enables scene reconstruction while novel view synthesis allows rendering of captured scenes from new viewpoints. However, both techniques face challenges with casual capture and dynamic scenes. To address this, a deferred approach that combines structure from motion and novel view synthesis is proposed, along with a variational diffusion formulation for enforcing spatio-temporal consistency.

COMPUTERS & GRAPHICS-UK (2022)

Add to Collection

Article Computer Science, Artificial Intelligence

MILO: Multi-Bounce Inverse Rendering for Indoor Scene With Light-Emitting Objects

Bohan Yu, Siqi Yang, Xuanning Cui, Siyan Dong, Baoquan Chen, Boxin Shi

Summary: In this paper, we propose a multi-bounce inverse rendering method based on Monte Carlo path tracing, which enables correct rendering of complex multi-bounce lighting effects in scene editing. We introduce a novel light source model suitable for indoor scenes, and design a specific neural network with disambiguation constraints to alleviate ambiguities during the inverse rendering. We evaluate our method on both synthetic and real indoor scenes through tasks such as virtual object insertion, material editing, and relighting, and the results demonstrate better photo-realistic quality.

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE (2023)

Add to Collection

Article Computer Science, Software Engineering

Fast and Accurate Illumination Estimation Using LDR Panoramic Images for Realistic Rendering

Haojie Cheng, Chunxiao Xu, Jiajun Wang, Zhenxin Chen, Lingxiao Zhao

Summary: This paper proposes an efficient method for accurately inferring illuminations of real-world scenes using LDR panoramic images. The method estimates lighting parameters such as light source locations, types, and intensities, and calculates realistic illumination intensities by recovering the inverse camera response function and extending the dynamic range of pixel values.

IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS (2023)

Add to Collection

Article Computer Science, Information Systems

Auto Complementary Exposure Control for High Dynamic Range Video Capturing

Bing Han, Xiu Jia, Rui Song, Feng Ran, Peng Rao

Summary: An auto complementary exposure (ACE) algorithm is proposed to maximize the entropy of HDR images, with advantages such as not requiring hardware modification, low computational load, and applicability to video recording.

IEEE ACCESS (2021)

Add to Collection

Article Computer Science, Artificial Intelligence

Photo-realistic image synthesis from lines and appearance with modular modulation

Wuyang Luo, Su Yang, Weishan Zhang

Summary: This paper proposes an image translation system driven by line and appearance images, which effectively combines different conditional information in image synthesis tasks. The structural and appearance features are progressively fused through cascaded modulation units, resulting in higher-quality transformation results.

NEUROCOMPUTING (2022)

Add to Collection

Article Computer Science, Artificial Intelligence

Film-GAN: towards realistic analog film photo generation

Haoyan Gong, Jionglong Su, Kah Phooi Seng, Anh Nguyen, Ao Liu, Hongbin Liu

Summary: This paper introduces Film-GAN, the first Generative Adversarial Network (GAN)-based method for translating digital images to film. By implementing the color-noise-encoding (CNE) network, Film-GAN outperforms other state-of-the-art approaches in terms of realism and image quality.

NEURAL COMPUTING & APPLICATIONS (2023)

Add to Collection

Article Computer Science, Software Engineering

Video Folding: Increased Framerate for Semi-Repetitive Sequences

Chris May, Manuel M. Oliveira, Daniel Aliaga

Summary: The technique introduced in the study seeks to increase the framerate of semi-repetitive videos by reordering and combining frames, resulting in a single non-repetitive sequence with higher temporal resolution. A novel frame warping technique is used to counteract differences between repetitions, maintaining smooth motion and enabling seamless, infinite looping. The effectiveness of the solution is demonstrated both quantitatively and qualitatively in the article and accompanying video.

IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS (2021)

Add to Collection

No Data Available

Article Computer Science, Software Engineering

Pervasive Augmented Reality to support real-time data monitoring in industrial scenarios: Shop floor visualization evaluation and user study

Rafael Maio, Tiago Araujo, Bernardo Marques, Andre Santos, Pedro Ramalho, Duarte Almeida, Paulo Dias, Beatriz Sousa Santos

Summary: Augmented Reality (AR) is a crucial technology in Industry 4.0 and smart manufacturing, particularly in the field of data monitoring. In this study, we developed a Pervasive AR tool for data monitoring, along with a web application for comparison purposes. User studies were conducted to gather feedback and evaluate the effectiveness of the systems, confirming the potential of Pervasive AR for data monitoring.

COMPUTERS & GRAPHICS-UK (2024)

Add to Collection

Article Computer Science, Software Engineering

Gaze-directed and saliency-guided approaches of stereo camera control in interactive virtual reality

Berk Cebeci, Mehmet Bahadir Askin, Tolga K. Capin, Ufuk Celikcan

Summary: Despite advances in virtual reality technologies, extended VR sessions with head-mounted displays (HMDs) still face challenges in terms of comfort. In this study, a methodology using gaze-directed and visual saliency-guided paradigms for automatic stereo camera control in real-time interactive VR was proposed. The results showed that the gaze-directed approach outperformed the saliency-guided approach, both improving the overall depth feeling without hindering visual comfort in the tested virtual environments (VEs).

COMPUTERS & GRAPHICS-UK (2024)

Add to Collection

Article Computer Science, Software Engineering

NOVAction23: Addressing the data diversity gap by uniquely generated synthetic sequences for real-world human action recognition

Ali Egemen Tasoren, Ufuk Celikcan

Summary: By developing the NOVAction engine, we have created the NOVAction23 dataset, which consists of highly diversified and photorealistic synthetic human action sequences. This dataset is significant in improving the performance of human action recognition.

COMPUTERS & GRAPHICS-UK (2024)

Add to Collection

© Peeref 2019-2024. All rights reserved.