Image inpainting based on deep learning has been greatly improved.The original purpose of image inpainting was to repair some broken photos, suchas inpainting artifacts. However, it may also be used for malicious oper...Image inpainting based on deep learning has been greatly improved.The original purpose of image inpainting was to repair some broken photos, suchas inpainting artifacts. However, it may also be used for malicious operations,such as destroying evidence. Therefore, detection and localization of imageinpainting operations are essential. Recent research shows that high-pass filteringfull convolutional network (HPFCN) is applied to image inpainting detection andachieves good results. However, those methods did not consider the spatial location and channel information of the feature map. To solve these shortcomings, weintroduce the squeezed excitation blocks (SE) and propose a high-pass filter attention full convolutional network (HPACN). In feature extraction, we apply concurrent spatial and channel attention (scSE) to enhance feature extraction and obtainmore information. Channel attention (cSE) is introduced in upsampling toenhance detection and localization. The experimental results show that the proposed method can achieve improvement on ImageNet.展开更多
As digital image techniques have been widely used, the requirements for high-resolution images become increasingly stringent. Traditional single-frame interpolation techniques cannot add new high frequency information...As digital image techniques have been widely used, the requirements for high-resolution images become increasingly stringent. Traditional single-frame interpolation techniques cannot add new high frequency information to the expanded images, and cannot improve resolution in deed. Multiframe-based techniques are effective ways for high-resolution image reconstruction, but their computation complexities and the difficulties in achieving image sequences limit their applications. An original method using an artificial neural network is proposed in this paper. Using the inherent merits in neural network, we can establish the mapping between high frequency components in low-resolution images and high-resolution images. Example applications and their results demonstrated the images reconstructed by our method are aesthetically and quantitatively (using the criteria of MSE and MAE) superior to the images acquired by common methods. Even for infrared images this method can give satisfactory results with high definition. In addition, a single-layer linear neural network is used in this paper, the computational complexity is very low, and this method can be realized in real time.展开更多
A novel design of high-efficiency broadband power amplifier （BPA） with the low-pass bias networkto enhance the efficiency and output power is presented in this paper. Compared with other bias networks, the proposed ...A novel design of high-efficiency broadband power amplifier （BPA） with the low-pass bias networkto enhance the efficiency and output power is presented in this paper. Compared with other bias networks, the proposed low-pass bias network shows a smaller baseband impedance, which can reduce the electrical memory effect. While it provides a larger radio frequency （RF） impedance, which can prevent the leakage of the output power from bias network. A BPA with the proposed bias network is designed using commercial GaN device Cree40025F. The designed BPA shows a fractional bandwidth of 40%, from 1.8 GHz to 2.7 GHz. The measured results exhibit 73.9 % drain efficiency （DE） value with output power of 43.5 dBm at 2.7 GHz, which appears an enhancement of 9.5% and 2.5 dBm comparing with that adopts LC bias network.展开更多
A C-band 6-bit digital phase shifter is presented. The phase shifter is based on the synthetic design of a high-pass/low-pass network and the all-pass network. The series scatter restrain method is also discussed. The...A C-band 6-bit digital phase shifter is presented. The phase shifter is based on the synthetic design of a high-pass/low-pass network and the all-pass network. The series scatter restrain method is also discussed. The phase shifter is fabricated in 0.25/zm GaAs PHEMT technology and developed for C-band phased arrays, and the relative phase shift varies from 0 to 360 in step of 5.625°. The phase shifter, with a chip size of 4 × 1.95 mm2, has achieved an insertion loss better than 6.4 dB, RMS phase error of less than 1.73°, and an input and output VSWR less than 1.6 at all conditions.展开更多
A low voltage,highly linear transconductance-C（G_m-C） low-pass filter for wireless local area network （WLAN） transceiver application is proposed.This transmitter（Tx） filter adopts a 9.8 MHz 3rd-order Chebyshev l...A low voltage,highly linear transconductance-C（G_m-C） low-pass filter for wireless local area network （WLAN） transceiver application is proposed.This transmitter（Tx） filter adopts a 9.8 MHz 3rd-order Chebyshev low pass prototype and achieves 35 dB stop-band attenuation at 30 MHz frequency.By utilizing pseudo-differential linear-region MOS transconductors,the filter IIP_3 is measured to be as high as 9.5 dBm.Fabricated in a 0.35μm standard CMOS technology,the proposed filter chip occupies a 0.41×0.17 mm^2 die area and consumes 3.36 mA from a 3.3-V power supply.展开更多
基金supported by the National Natural Science Foundation of China under Grant 62172059,61972057 and 62072055Hunan Provincial Natural Science Foundations of China under Grant 2020JJ4626+1 种基金Scientific Research Fund of Hunan Provincial Education Department of China under Grant 19B004Postgraduate Scientific Research Innovation Project of Hunan Province under Grant CX20210811.
文摘Image inpainting based on deep learning has been greatly improved.The original purpose of image inpainting was to repair some broken photos, suchas inpainting artifacts. However, it may also be used for malicious operations,such as destroying evidence. Therefore, detection and localization of imageinpainting operations are essential. Recent research shows that high-pass filteringfull convolutional network (HPFCN) is applied to image inpainting detection andachieves good results. However, those methods did not consider the spatial location and channel information of the feature map. To solve these shortcomings, weintroduce the squeezed excitation blocks (SE) and propose a high-pass filter attention full convolutional network (HPACN). In feature extraction, we apply concurrent spatial and channel attention (scSE) to enhance feature extraction and obtainmore information. Channel attention (cSE) is introduced in upsampling toenhance detection and localization. The experimental results show that the proposed method can achieve improvement on ImageNet.
文摘As digital image techniques have been widely used, the requirements for high-resolution images become increasingly stringent. Traditional single-frame interpolation techniques cannot add new high frequency information to the expanded images, and cannot improve resolution in deed. Multiframe-based techniques are effective ways for high-resolution image reconstruction, but their computation complexities and the difficulties in achieving image sequences limit their applications. An original method using an artificial neural network is proposed in this paper. Using the inherent merits in neural network, we can establish the mapping between high frequency components in low-resolution images and high-resolution images. Example applications and their results demonstrated the images reconstructed by our method are aesthetically and quantitatively (using the criteria of MSE and MAE) superior to the images acquired by common methods. Even for infrared images this method can give satisfactory results with high definition. In addition, a single-layer linear neural network is used in this paper, the computational complexity is very low, and this method can be realized in real time.
基金supported by National Basic Research Program of China(973 Program)(2014CB339900)National Natural Science Foundation of China(61201025)National Natural Science Foundation of China for the Major Equipment Development(61327806)
文摘A novel design of high-efficiency broadband power amplifier （BPA） with the low-pass bias networkto enhance the efficiency and output power is presented in this paper. Compared with other bias networks, the proposed low-pass bias network shows a smaller baseband impedance, which can reduce the electrical memory effect. While it provides a larger radio frequency （RF） impedance, which can prevent the leakage of the output power from bias network. A BPA with the proposed bias network is designed using commercial GaN device Cree40025F. The designed BPA shows a fractional bandwidth of 40%, from 1.8 GHz to 2.7 GHz. The measured results exhibit 73.9 % drain efficiency （DE） value with output power of 43.5 dBm at 2.7 GHz, which appears an enhancement of 9.5% and 2.5 dBm comparing with that adopts LC bias network.
文摘A C-band 6-bit digital phase shifter is presented. The phase shifter is based on the synthetic design of a high-pass/low-pass network and the all-pass network. The series scatter restrain method is also discussed. The phase shifter is fabricated in 0.25/zm GaAs PHEMT technology and developed for C-band phased arrays, and the relative phase shift varies from 0 to 360 in step of 5.625°. The phase shifter, with a chip size of 4 × 1.95 mm2, has achieved an insertion loss better than 6.4 dB, RMS phase error of less than 1.73°, and an input and output VSWR less than 1.6 at all conditions.
文摘A low voltage,highly linear transconductance-C（G_m-C） low-pass filter for wireless local area network （WLAN） transceiver application is proposed.This transmitter（Tx） filter adopts a 9.8 MHz 3rd-order Chebyshev low pass prototype and achieves 35 dB stop-band attenuation at 30 MHz frequency.By utilizing pseudo-differential linear-region MOS transconductors,the filter IIP_3 is measured to be as high as 9.5 dBm.Fabricated in a 0.35μm standard CMOS technology,the proposed filter chip occupies a 0.41×0.17 mm^2 die area and consumes 3.36 mA from a 3.3-V power supply.
