Deep learning has achieved remarkable results in the field of target detection and recognition. For small targets in images, image pyramid can be used to fuse multi-scale features to improve detection performance. How...
详细信息
To benefit network transmission, the bit stream of the whole frame coded by H.264/AVC is usually grouped into one packet. However, the packet loss during transmission will lead to the distortion of the reconstructed v...
详细信息
Multi-object tracking is important in many computer vision applications. The major difficulties are due to inter-object or scene occlusion and data association. In this paper, we present a method to automatically dete...
详细信息
A watermarking scheme designed for remote sensing images needs to meet the same demand of both invisibility as for ordinary digital images. Due to specific perceptual characteristics of Synthetic Aperture Radar(SAR) i...
详细信息
A watermarking scheme designed for remote sensing images needs to meet the same demand of both invisibility as for ordinary digital images. Due to specific perceptual characteristics of Synthetic Aperture Radar(SAR) images, the watermarking algorithms with consideration of Human Vision system(HVS) modeling from optical images give poor performance when applied on SAR images. This paper examines a variety of factors affecting the noise sensitivity, and further proposes a refined pixel-wise masking approach for watermarking on SAR images. The proposed approach is applied on logarithmic transformed SAR images, and has increased the acceptable watermark embedding strength by about 6 dB to 10 dB while achieving the same levels of watermarked image visual quality. Experimental results show that this approach enhanced the perceptual invisibility of watermarking based on wavelet decomposition.
We study the problem of recognizing sign language automatically using the RGB videos and skeleton coordinates captured by Kinect, which is of great significance in communication between the deaf and the hearing societ...
详细信息
In many image-related tasks, learning expressive and discriminative representations of images is essential, and deep learning has been studied for automating the learning of such representations. Some user-centric tas...
详细信息
ISBN:
(纸本)9781467388511
In many image-related tasks, learning expressive and discriminative representations of images is essential, and deep learning has been studied for automating the learning of such representations. Some user-centric tasks, such as image recommendations, call for effective representations of not only images but also preferences and intents of users over images. Such representations are termed hybrid and addressed via a deep learning approach in this paper. We design a dual-net deep network, in which the two subnetworks map input images and preferences of users into a same latent semantic space, and then the distances between images and users in the latent space are calculated to make decisions. We further propose a comparative deep learning (CDL) method to train the deep network, using a pair of images compared against one user to learn the pattern of their relative distances. The CDL embraces much more training data than naive deep learning, and thus achieves superior performance than the latter, with no cost of increasing network complexity. Experimental results with real-world data sets for image recommendations have shown the proposed dual-net network and CDL greatly outperform other stateof-the-art image recommendation solutions.
There exit high variations among nano-devices in nano-electronic systems, owing to the extremely small size and the bottom-up self-assembly nanofabrication process. Therefore, it is important to develop logical functi...
详细信息
ISBN:
(纸本)9781450328814
There exit high variations among nano-devices in nano-electronic systems, owing to the extremely small size and the bottom-up self-assembly nanofabrication process. Therefore, it is important to develop logical function mapping techniques with the consideration of variation tolerance. In this paper, the variation tolerant logical mapping (VTLM) problem is treated as a multiobjective optimization problem (MOP), a hybridization of Nondominated Sorting Genetic Algorithm II (NSGA-II) with a problem-specific local search is presented to solve the problem. The experiment results show that with the assistance of the problem-specific local search, the presented algorithm is effective, and can find better solutions than that without the local search.
Sparse synthetic aperture radar(SAR) imaging has emerged as a reliable microwave imaging scheme in the recent decade and excels in down-sampling reconstruction and full-sampling performance improvements such as noise,...
详细信息
Sparse synthetic aperture radar(SAR) imaging has emerged as a reliable microwave imaging scheme in the recent decade and excels in down-sampling reconstruction and full-sampling performance improvements such as noise, sidelobe, speckle, and ambiguity suppression. To utilize complex image products of sparse reconstruction for improvement in polarimetric, interferometric, and tomographic SAR imaging, it is necessary to evaluate the phase preservation of sparse SAR imaging. In this study, we first introduce the general alternating direction method of multipliers(ADMM) as the universal framework for sparse reconstruction algorithms and adopt chirp scaling algorithm(CSA)-based azimuth-range decouple operators to avoid expensive data storage and processing. Further, we theoretically analyze the phase preservation of the sparse reconstruction algorithm through a comparison with the reconstruction results of CSA. Finally,we conduct the interferometric offset test on the sparse reconstruction results of simulated and real Gaofen-3(GF-3) SAR data, demonstrating the phase-preserving ability of sparse methods.
Currently, many studies use Fourier amplitude spectra of speech signals to predict depression levels. However, those works often treat Fourier amplitude spectra as images or sequences to capture depression cues using ...
详细信息
High resolution is a key trend in the development of synthetic aperture radar (SAR), which enables the capture of fine details and accurate representation of backscattering properties. However, traditional high-resolu...
详细信息
暂无评论