4.7 Article

Design of a Low-Power Super-Resolution Architecture for Virtual Reality Wearable Devices

相关参考文献

注意:仅列出部分参考文献,下载原文获取全部文献信息。
Article Computer Science, Hardware & Architecture

ADAS: A High Computational Utilization Dynamic Reconfigurable Hardware Accelerator for Super Resolution

Liang Chang et al.

Summary: This study proposes an all-on-chip hardware architecture called ADAS, which utilizes the deconvolution scheme and feature map segmentation strategy to improve hardware implementation efficiency in super-resolution image reconstruction. ADAS achieves real-time image SR and better image reconstruction quality with 37.15dB PSNR and 0.9587 SSIM. It also supports multiple scaling factors and achieves significant speedup compared to the baseline.

ACM TRANSACTIONS ON RECONFIGURABLE TECHNOLOGY AND SYSTEMS (2023)

Article Geochemistry & Geophysics

Satellite Video Super-Resolution via Multiscale Deformable Convolution Alignment and Temporal Grouping Projection

Yi Xiao et al.

Summary: A novel fusion strategy of temporal grouping projection and an accurate alignment module is proposed for satellite video super-resolution (VSR) in this article. By enhancing the alignment and fusion methods, the spatial resolution and dynamic analysis of satellite videos are effectively improved. Extensive experiments demonstrate that this method outperforms current state-of-the-art VSR methods on Jilin-1 satellite video.

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING (2022)

Article Computer Science, Artificial Intelligence

Real-world super-resolution of face-images from surveillance cameras

Andreas Aakerberg et al.

Summary: This study introduces a novel framework for generating realistic LR/HR training pairs to address the issue of poor performance of existing SR methods on real LR images. Experimental results demonstrate that the proposed method leads to more detailed reconstructions and less noise.

IET IMAGE PROCESSING (2022)

Article Engineering, Electrical & Electronic

An Attention Encoder-Decoder Network Based on Generative Adversarial Network for Remote Sensing Image Dehazing

Liquan Zhao et al.

Summary: This article proposes an encoder-decoder method based on generative adversarial network to address the problem of remote sensing image dehazing. The method learns image features using low-frequency and high-frequency information, and incorporates skip connections, multi-scale attention module, CBlock module, and distillation module to enhance the dehazing ability of the network. Experimental results demonstrate that this method achieves the best performance on the RICE dataset, both qualitatively and quantitatively.

IEEE SENSORS JOURNAL (2022)

Article Engineering, Electrical & Electronic

Video Super-Resolution for Wireless Capsule Endoscopy Imaging Sensor

Chao Zhou et al.

Summary: This study addresses two challenges in wireless capsule endoscopy (WCE) applications, namely the lack of high-resolution counterparts and the difficulty of utilizing temporal correlations to enhance recovery performance. A novel training dataset and an effective alignment network are proposed to tackle these challenges. Experimental results demonstrate the superiority of the proposed method in practical WCE applications over other approaches.

IEEE SENSORS JOURNAL (2022)

Proceedings Paper Computer Science, Information Systems

Super Resolution for Augmented Reality Applications

Vladislav Li et al.

Summary: This paper discusses integrating new object detection and classification technologies using super-resolution technology into small AR/VR devices, aiming to propose a low-footprint Generative Adversarial Network (GAN) framework for SR-supported recognition model.

IEEE INFOCOM 2022 - IEEE CONFERENCE ON COMPUTER COMMUNICATIONS WORKSHOPS (INFOCOM WKSHPS) (2022)

Proceedings Paper Engineering, Electrical & Electronic

BSRA: Block-based Super Resolution Accelerator with Hardware Efficient Pixel Attention

Dun-Hao Yang et al.

Summary: This paper proposes a super resolution hardware accelerator with hardware efficient pixel attention that achieves better reconstruction results. It adopts full model block wise convolution and PE arrays with distributed weights to support CNN and pixel attention. The final implementation can perform full HD image reconstruction at 30 frames per second.

2022 IEEE INTERNATIONAL SYMPOSIUM ON CIRCUITS AND SYSTEMS (ISCAS 22) (2022)

Proceedings Paper Computer Science, Cybernetics

RA360SR: A Real-time Acceleration-adaptive 360-degree Video Super-resolution System

Jiapeng Chi et al.

Summary: This paper presents RA360SR, a real-time acceleration-adaptive 360-degree video super-resolution system. By using a dual-camera system and Unity3D post-processing, it delivers a sharper video viewing experience. The system employs an acceleration-adaptive approach to switch the super-resolution model processing status based on the acceleration of users' head movements, ensuring a more stable frame rate.

2022 IEEE INTERNATIONAL SYMPOSIUM ON MIXED AND AUGMENTED REALITY ADJUNCT (ISMAR-ADJUNCT 2022) (2022)

Article Chemistry, Multidisciplinary

Recent Advances in Wearable Sensors and Integrated Functional Devices for Virtual and Augmented Reality Applications

Hojoong Kim et al.

Summary: This report highlights the recent advancements in multifunctional wearable sensors and integrated functional devices for VR/AR applications, emphasizing the importance of enhanced sensor functions and seamless integration for more natural interactions with the virtual world.

ADVANCED FUNCTIONAL MATERIALS (2021)

Proceedings Paper Computer Science, Artificial Intelligence

COMISR: Compression-Informed Video Super-Resolution

Yinxiao Li et al.

Summary: The paper proposes a compression-informed video super-resolution model which includes three modules to handle compression properties, allowing for high-resolution content restoration across various compression rates and achieving state-of-the-art performance in multiple metrics.

2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021) (2021)

Article Computer Science, Information Systems

360° Image Reference-Based Super-Resolution Using Latitude-Aware Convolution Learned From Synthetic to Real

Hee-Jae Kim et al.

Summary: This paper presents an efficient reference-based 360-degree image super-resolution technique that utilizes the wide field of view between adjacent 360-degree cameras to enhance the resolution. Various methods, such as DE360 and LatConv, are used to overcome distortion and maintain image quality. By using synthetic image datasets and a synthetic-to-real learning scheme, knowledge learned from synthetic images is successfully transferred to real camera-captured images.

IEEE ACCESS (2021)

Article Imaging Science & Photographic Technology

Design of Flexible Hardware Accelerators for Image Convolutions and Transposed Convolutions

Cristian Sestito et al.

Summary: With the increasing real-time and low-power requirements in modern computer vision applications, researchers are exploring the deployment of CNNs on hardware platforms such as FPGAs. This paper proposes a novel hardware-oriented algorithm to efficiently accelerate both CONVs and TCONVs, resulting in improved computational efficiency.

JOURNAL OF IMAGING (2021)

Article Engineering, Electrical & Electronic

An Energy-Efficient FPGA-Based Deconvolutional Neural Networks Accelerator for Single Image Super-Resolution

Jung-Woo Chang et al.

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY (2020)

Article Engineering, Electrical & Electronic

SRNPU: An Energy-Efficient CNN-Based Super-Resolution Processor With Tile-Based Selective Super-Resolution in Mobile Devices

Juhyoung Lee et al.

IEEE JOURNAL ON EMERGING AND SELECTED TOPICS IN CIRCUITS AND SYSTEMS (2020)

Editorial Material Engineering, Electrical & Electronic

Energy-Quality Scalable Circuits and Systems for Sensing and Computing: From Approximate to Communication-Inspired and Learning-Based

Massimo Alioto et al.

IEEE JOURNAL ON EMERGING AND SELECTED TOPICS IN CIRCUITS AND SYSTEMS (2018)

Proceedings Paper Engineering, Electrical & Electronic

Accurate and Bandwidth Efficient Architecture for CNN-based Full-HD Super-Resolution

Chung-Yan Chih et al.

2018 IEEE INTERNATIONAL SYMPOSIUM ON CIRCUITS AND SYSTEMS (ISCAS) (2018)

Article Computer Science, Hardware & Architecture

Scaling equations for the accurate prediction of CMOS device performance from 180 nm to 7 nm

Aaron Stillmaker et al.

INTEGRATION-THE VLSI JOURNAL (2017)

Article Computer Science, Software Engineering

Latency Requirements for Foveated Rendering in Virtual Reality

Rachel Albert et al.

ACM TRANSACTIONS ON APPLIED PERCEPTION (2017)

Article Computer Science, Software Engineering

Towards Foveated Rendering for Gaze-Tracked Virtual Reality

Anjul Patney et al.

ACM TRANSACTIONS ON GRAPHICS (2016)

Article Engineering, Electrical & Electronic

Super-resolution image reconstruction: A technical overview

SC Park et al.

IEEE SIGNAL PROCESSING MAGAZINE (2003)