We propose a deep convolutional neural network architecture codenamed Inception that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14). The main hallmark of this architecture is the improved utilization of the computing resources inside the network. By a carefully crafted design, we increased the depth and width of the network while keeping the computational budget constant. To optimize quality, the architectural decisions were based on the Hebbian principle and the intuition of multi-scale processing. One particular incarnation used in our submission for ILSVRC14 is called GoogLeNet, a 22 layers deep network, the quality of which is assessed in the context of classification and detection.

/pdf/going-deeper-with-convolutions-1yobw2o2ds.pdf

Going deeper with convolutions

[신간의 별자리x] 우리/미술, 그리고 ‘슬픔의 박물관’

Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

I have developed "tennis elbow" from lugging this book around the past four weeks, but it is worth the pain, the effort, and the aspirin. It is also worth the (relatively speaking) bargain price. Including appendixes, this book contains 894 pages of text. The entire panorama of the neural sciences is surveyed and examined, and it is comprehensive in its scope, from genomes to social behaviors. The editors explicitly state that the book is designed as "an introductory text for students of biology, behavior, and medicine," but it is hard to imagine any audience, interested in any fragment of neuroscience at any level of sophistication, that would not enjoy this book. The editors have done a masterful job of weaving together the biologic, the behavioral, and the clinical sciences into a single tapestry in which everyone from the molecular biologist to the practicing psychiatrist can find and appreciate his or

Principles of Neural Science

Despite the breakthroughs in accuracy and speed of single image super-resolution using faster and deeper convolutional neural networks, one central problem remains largely unsolved: how do we recover the finer texture details when we super-resolve at large upscaling factors? The behavior of optimization-based super-resolution methods is principally driven by the choice of the objective function. Recent work has largely focused on minimizing the mean squared reconstruction error. The resulting estimates have high peak signal-to-noise ratios, but they are often lacking high-frequency details and are perceptually unsatisfying in the sense that they fail to match the fidelity expected at the higher resolution. In this paper, we present SRGAN, a generative adversarial network (GAN) for image super-resolution (SR). To our knowledge, it is the first framework capable of inferring photo-realistic natural images for 4x upscaling factors. To achieve this, we propose a perceptual loss function which consists of an adversarial loss and a content loss. The adversarial loss pushes our solution to the natural image manifold using a discriminator network that is trained to differentiate between the super-resolved images and original photo-realistic images. In addition, we use a content loss motivated by perceptual similarity instead of similarity in pixel space. Our deep residual network is able to recover photo-realistic textures from heavily downsampled images on public benchmarks. An extensive mean-opinion-score (MOS) test shows hugely significant gains in perceptual quality using SRGAN. The MOS scores obtained with SRGAN are closer to those of the original high-resolution images than to those obtained with any state-of-the-art method.

/pdf/photo-realistic-single-image-super-resolution-using-a-2y0438rlmd.pdf

Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network

Separating the background and foreground components from video frames is important to many tasks in computer vision and multimedia. As of today, robust principal component analysis (RPCA) has shown highly promising performance with the assumption that the background is low-rank and the foreground is sparse. However, existing RPCA-based methods have overlooked the uncertainty that some parts of the background (e.g., moving leaves in a dynamic background) or even the whole background (e.g., camera jittering) can be moving, which violates the low-rank assumption. To address this issue, we propose a novel enhanced RPCA framework (called ERPCA) by robustly modeling the dynamic background. Different from traditional RPCA framework, the background is decomposed into a low-rank component and a sparse component in the proposed ERPCA framework. Specifically, the sparse parts including foreground and dynamic parts of the background are modeled by Gaussian scale mixture (GSM) model. Moreover, those sparse components are further constrained by temporal consistency using nonzeromeans Gaussian models; the correspondences between sparse pixels in adjacent frames are explored by optical flow. Experimental results on 40 real videos demonstrate the superiority of our proposed method, with better average results than current state-of-the-art foreground estimation methods.

Robust Dynamic Background Modeling for Foreground Estimation

The target of high resolution digital terrain elevation sequence is to obtain 3-D representation of a terrain's surface. In this work we used interferometric technique which is based on the analysis of the phase difference between the signals received at two antenna position. Under suitable conditions, the resulting DEM of this method is locally the most precise one. This work is concerned with the exploitation of radar SLC images of the Las Vega area acquired on 24 May and 25 May 1994, the first is taken as the master image and the other as slave. We have developed an interferometric process beginning with the implementation of a co-registration algorithm for the images. Then we developed the flatearth phase elimination from the interferogram. We also developed a filter in order to minimiz the existing noise. At the end remove the effect of under sampling to get a good result in the final step of the process which is the unwrapping step.

https://asat.journals.ekb.eg/article_22190_69315fbceb9d73df36f7dcb8c81de382.pdf

High Resolution Digital Terrain Elevation Processing

Single-sensor digital video cameras use a color filter array (CFA) to capture video and a color demosaicking (CDM) procedure to reproduce the full color sequence. The reproduced video frames suffer from the inevitable sensor noise introduced in the video acquisition process. This paper presents a spatial-temporal denoising and demosaicking scheme that works without explicit motion estimation. We first perform patch based denoising on the mosaic CFA video. For each CFA patch to be denoised, similar patches are selected within a local spatial- temporal neighborhood. The principal component analysis is performed on the selected patches to remove noise. We then apply an initial single-frame CDM to the denoised CFA data, and subsequently post-process the demosaicked frames by exploiting the spatial-temporal redundancy to reduce the color artifacts. The experimental results on simulated and real noisy CFA sequences demonstrate that the proposed spatial-temporal CFA video de- noising and demosaicking scheme can significantly reduce the noise-caused color artifacts and effectively preserve the image edge structures. Index Terms—Color demosaicking (CDM), color filter array (CFA), denoising, video enhancement.

Spatial-Temporal Color Video Reconstruction From

This paper presents a joint rate-distortion optimization coding scheme for H.264/AVC intra coding, which uses cluster computing technologies to jointly optimize coding parameters among different image blocks. We identify two kinds of dependencies, inter sub-block dependency and inter macroblock dependency, ignored in the ordinary rate-distortion optimization and a joint RD cost is designed in the optimization process to consider those dependencies. We show that with our proposed scheme, a significant better coding performance can be achieved with the standard-compliant bitstreams. Experiment results show up to 5% bits saving compared to the H.264/AVC reference software.

Joint rate-distortion optimization for H.264/AVC intra coding based on cluster computing

Spectral video is crucial for monitoring of dynamic scenes, reconnaissance of moving targets, observation and tracking 
of living cells, etc. The traditional spectral imaging methods need multiple exposures to capture a full frame spectral 
image, which leads to a low temporal resolution and thus lose their value as spectral video. The new code aperture 
snapshot spectral imaging (CASSI) method has been emerging in recent years, which is suitable for spectral video 
acquisition, due to its high-speed snapshot and few-amount measurements. Based on the CASSI, this paper proposes a 
compressive spectral video acquisition method with double-channel complementary coded aperture. The method can 
achieve the spectral video with a high temporal resolution by directly sampling the 3D spectral scene with 2D array 
sensor in only one snapshot. Furthermore, by using the double-channel complementary coded aperture in compressive 
measurement and the sparse regularity in the optimization recovery together, we can obtain the higher PSNR and better 
visual effects compared with the single-channel CASSI. Simulation results demonstrate the efficacy of the proposed 
method.

Guangming Shi

Papers

Robust Dynamic Background Modeling for Foreground Estimation

High Resolution Digital Terrain Elevation Processing

Spatial-Temporal Color Video Reconstruction From

Joint rate-distortion optimization for H.264/AVC intra coding based on cluster computing

Compressive Spectral Video Acquisition with Double-channel Complementary Coded Aperture