Article
Engineering, Electrical & Electronic
Chuanmin Jia, Xinyu Hang, Shanshe Wang, Yaqiang Wu, Siwei Ma, Wen Gao
Summary: The recent trend in neural image compression research can be classified into two categories: improvements in analysis-synthesis transform networks and optimization of entropy estimation. By utilizing more expressive network structures and advanced entropy models, NIC compression efficiency is enhanced. Additionally, expanding the horizon of NIC to hardware-based compression on resource-constrained platforms has been explored. A novel FPGA-accelerated NIC framework has been proposed for hardware encoding, enabling real-time edge-computing video services.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2022)
Article
Engineering, Electrical & Electronic
Yang Zhu, Sen Wang, Yinhui Zhang, Zifen He, Qingjian Wang
Summary: In this article, a structural body image super-resolution network based on a dual transformer architecture is proposed to improve the clarity of the collected structural body vibration displacement image and better capture the vibration displacement information of the target. By introducing an attention mechanism and interacting with feature information in different layers of the encoder-decoder architecture, a better structural body image super-resolution effect is achieved.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Article
Engineering, Electrical & Electronic
Yichen Chi, Wenming Yang, Yapeng Tian
Summary: In this study, a method for restoring ultra-high-resolution images is proposed. By using a global degradation supervision super-resolution (GDSSR) method, this approach is able to restore the local patches of the image and maintain consistency among them. Experimental results demonstrate that the GDSSR method achieves superior restoration performance on ultra-high-resolution image datasets.
IEEE SIGNAL PROCESSING LETTERS
(2023)
Article
Engineering, Electrical & Electronic
Yoshiaki Shishikui, Yasuhito Sawahata
Summary: This study aimed to investigate the psychological effects of 8K ultra-high-definition television (UHDTV) under daily TV viewing conditions. The findings suggest that increasing resolution from 4K to 8K enhances impressions even at viewing distances beyond the design viewing distance, confirming strong impressions provided by 8K across a wide viewing range. Additionally, the study derived the relationship between higher-order impressions and lower-order impressions, providing insights for production and post-processing.
IEEE TRANSACTIONS ON BROADCASTING
(2022)
Article
Chemistry, Multidisciplinary
Yuan Qin, Li-Heng Li, Zhaoan Yu, Feihong Wu, Danian Dong, Wei Guo, Zhongfang Zhang, Jun-Hui Yuan, Kan-Hao Xue, Xiangshui Miao, Shibing Long
Summary: This study focuses on ultrahigh-performance metal-semiconductor-metal (MSM) solar-blind photodetectors (SBPDs) based on post-annealed amorphous (a-) Ga2O3, demonstrating superhigh sensitivity and response speed, as well as ultrahigh photo-to-dark current ratio and specific detectivity, indicating practicality for applications in solar-blind imaging, environmental monitoring, artificial intelligence, and machine vision.
Article
Chemistry, Analytical
Zefeng Ying, Da Pan, Ping Shi
Summary: In this study, a no-reference VQA method was proposed to effectively and efficiently predict quality scores for UHD videos. A spatial distortion feature network based on a super-resolution model was constructed to quickly extract the global spatial distortion features of UHD videos. Then, a time fusion network based on a reinforcement learning model was used to aggregate the spatial distortion features of each UHD frame, achieving competitive quality prediction performance with a shorter runtime and fewer model parameters compared to other state-of-the-art VQA methods.
Article
Engineering, Electrical & Electronic
Yupeng Wang, Wenjie Jiang, Honglan Chen, Xinwei Chen, Yuan Gao, Pengfei Tian, Baoqing Sun
Summary: This method based on TGI enables signal transmission at an ultra-high frequency, surpassing the hardware bandwidth limit and achieving high quality and recognizability. It has potential applications in various fields of optical communication.
JOURNAL OF LIGHTWAVE TECHNOLOGY
(2022)
Article
Biochemical Research Methods
Qiuyuan Zhong, Anan Li, Rui Jin, Dejie Zhang, Xiangning Li, Xueyan Jia, Zhangheng Ding, Pan Luo, Can Zhou, Chenyu Jiang, Zhao Feng, Zhihong Zhang, Hui Gong, Jing Yuan, Qingming Luo
Summary: The development of LiMo technique and HD-fMOST method allows for high-throughput imaging of thick tissues and substantial improvement in neuronal morphology reconstruction. By achieving >30-fold lossless data compression at a voxel resolution of 0.32 x 0.32 x 1.00 mu m(3), real-time high-precision (95% accuracy) brain-wide 3D cell counting is made possible.
Article
Computer Science, Information Systems
Yunning Feng, Nan Hu, Xiaosheng Yu
Summary: This paper focuses on the optimization algorithm of ultra-high-definition live streaming and proposes an image superresolution algorithm based on hybrid dilated convolution and Laplacian pyramid. The experiment results demonstrate that the proposed algorithm outperforms baselines in terms of peak signal to noise ratio, structural similarity index measurement, and image quality.
WIRELESS COMMUNICATIONS & MOBILE COMPUTING
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Kaihao Zhang, Dongxu Li, Wenhan Luo, Wenqi Ren, Bjorn Stenger, Wei Liu, Hongdong Li, Ming-Hsuan Yang
Summary: This paper introduces two large-scale image datasets, UHDSR4K and UHDSR8K, to evaluate existing SISR methods on 4K and 8K resolution images. Additionally, a baseline model called MANet is proposed, utilizing an attention mechanism to learn correlations among feature maps and focus on more important features.
2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021)
(2021)
Article
Chemistry, Analytical
Venkatesh Kodukula, Saad Katrawala, Britton Jones, Carole-Jean Wu, Robert LiKamWa
Summary: Research shows that near-sensor processing can reduce system power but degrade image quality, requiring a balance between maintaining sensor temperature within an acceptable range. Temperature can be regulated based on dynamic visual task requirements to meet the needs of different application scenarios.
Article
Engineering, Civil
Hamed Haghighi, Mehrdad Dianati, Valentina Donzella, Kurt Debattista
Summary: Camera image simulation is crucial for virtual validation of autonomous vehicles and robots, as well as for creating image datasets for training vision models. To address the computational complexity, we propose a technique based on Stereo Super Resolution (SSR) to speed up the simulation of stereo images, achieving promising results in terms of speed and performance.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Computer Science, Information Systems
Xiao Tan, Huaian Chen, Kai Xu, Yi Jin, Changan Zhu
Summary: This paper proposes a deep neural network called Deep SR-HDR for the joint task of super-resolution and high dynamic range imaging. The network reconstructs high-resolution HDR images from a set of differently exposed low-resolution LDR images of a dynamic scene. By merging the shared processing steps and designing a multi-scale deformable module, the proposed network achieves high-quality image reconstruction efficiently.
IEEE TRANSACTIONS ON MULTIMEDIA
(2023)
Article
Engineering, Environmental
Limei Tian, Haixu Dou, Yanlong Shao, Yaozhen Yi, Xue Fu, Jie Zhao, Yong Fan, Weihua Ming, Luquan Ren
Summary: A silicone-oil-infused magnetic porous polydimethylsiloxane (SMPP) surface with magnetically tunable super-wetting behaviors is reported. This surface allows for precise control of droplet adhesion by manipulating the seepage and absorption of silicone oil using a magnetic field. It can be used for droplet manipulation, microfluidic microreactors, and antigravity transport devices.
CHEMICAL ENGINEERING JOURNAL
(2023)
Article
Computer Science, Artificial Intelligence
Rui Dou, Jiawen Li, Xujie Wan, Heyou Chang, Hao Zheng, Guangwei Gao
Summary: The article introduces an architecture called the Decoder Structure Guided CNN-Transformer Network (DCTNet) for face image super-resolution. DCTNet utilizes a decoder structure as its backbone, focusing primarily on Global-Local Feature Extraction Units (GLFEU).
IET COMPUTER VISION
(2023)
Article
Optics
Naoto Okaichi, Masato Miura, Hisayuki Sasaki, Hayato Watanabe, Jun Arai, Masahiro Kawakita, Tomoyuki Mishina
OPTICAL ENGINEERING
(2018)
Article
Optics
Masahiro Kawakita, Shoichro Iwasawa, Roberto Lopez-Gulliver, Naomi Inoue
OPTICAL ENGINEERING
(2018)
Article
Engineering, Electrical & Electronic
Naoto Okaichi, Masahiro Kawakita, Hisayuki Sasaki, Hayato Watanabe, Tomoyuki Mishina
JOURNAL OF THE SOCIETY FOR INFORMATION DISPLAY
(2019)
Article
Optics
Hisayuki Sasaki, Naoto Okaichi, Hayato Watanabe, Masanori Kano, Masato Miura, Masahiro Kawakita, Tomoyuki Mishina
Article
Multidisciplinary Sciences
Hayato Watanabe, Naoto Okaichi, Takuya Omura, Masanori Kano, Hisayuki Sasaki, Masahiro Kawakita
SCIENTIFIC REPORTS
(2019)
Article
Optics
Hayato Watanabe, Naoto Okaichi, Hisayuki Sasaki, Masahiro Kawakita
Article
Optics
Naoto Okaichi, Hisayuki Sasaki, Masanori Kano, Jun Arai, Masahiro Kawakita, Takeshi Naemura
Summary: In this study, an integral three-dimensional (3D) display system with a wide viewing zone and depth range was proposed using time-division display and eye-tracking technology. The results showed that the proposed system achieved a wider depth range for 3D images compared to other systems.
OPTICAL ENGINEERING
(2022)
Article
Optics
Takuya Omura, Hayato Watanabe, Masahiro Kawakita, Jun Arai
Summary: A flicker-reduced time-division light ray quadruplexing technology is proposed to improve the spatial and angular resolutions of 3D images. By optimizing the design of an image-shift optical device and using a 120 Hz 8K projector with a wobbling device and a wavelength-selective plate, high-resolution 3D image display with deeper depth can be achieved.
Article
Optics
Naoto Okaichi, Hisayuki Sasaki, Masanori Kano, Jun Arai, Masahiro Kawakita, Takeshi Naemura
Summary: A novel integral 3D display system is proposed in this study, utilizing eye-tracking technology and a lens array with a long focal length to enhance the viewing zone and depth range of reconstructed 3D images. A rotating lens arrangement is designed to reduce crosstalk caused by flipped images due to viewer movement, resulting in significantly expanded viewing angles and doubled depth range. Results from evaluation experiments confirmed effective suppression of crosstalk.
Proceedings Paper
Computer Science, Artificial Intelligence
Hisayuki Sasaki, Naoto Okaichi, Hayato Watanabe, Takuya Omura, Masanori Kano, Masahiro Kawakita
2020 IEEE INTERNATIONAL SYMPOSIUM ON MIXED AND AUGMENTED REALITY (ISMAR 2020)
(2020)
Proceedings Paper
Computer Science, Artificial Intelligence
Hisayuki Sasaki, Hayato Watanabe, Naoto Okaichi, Kensuke Hisatomi, Masahiro Kawakita
2019 26TH IEEE CONFERENCE ON VIRTUAL REALITY AND 3D USER INTERFACES (VR)
(2019)
Proceedings Paper
Optics
Hayato Watanabe, Naoto Okaichi, Takuya Omura, Masanori Kano, Hisayuki Sasaki, Masahiro Kawakita
THREE-DIMENSIONAL IMAGING, VISUALIZATION, AND DISPLAY 2019
(2019)
Proceedings Paper
Engineering, Electrical & Electronic
Kazuhiro Hara, Miwa Katayama, Masahiro Kawakita, Toshiaki Fujii, Tomoyuki Mishina
2018 ASIA-PACIFIC SIGNAL AND INFORMATION PROCESSING ASSOCIATION ANNUAL SUMMIT AND CONFERENCE (APSIPA ASC)
(2018)
Proceedings Paper
Optics
Hayato Watanabe, Masahiro Kawakita, Naoto Okaichi, Hisayuki Sasaki, Masanori Kano, Tomoyuki Mishina
THREE-DIMENSIONAL IMAGING, VISUALIZATION, AND DISPLAY 2018
(2018)
Proceedings Paper
Optics
M. Kawakita, H. Sasaki, N. Okaichi, H. Watanabe, M. Kano, J. Arai, T. Mishina
THREE-DIMENSIONAL IMAGING, VISUALIZATION, AND DISPLAY 2018
(2018)