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

Nonlocal center-surround reconstruction-based bottom-up saliency estimation

Total variation (TV) regularisation has been widely used for compressive sensing (CS) reconstruction. However, since TV regularisers favour piecewise constant solutions, they tend to produce over-smoothed image edges. To overcome this drawback, proposed is a novel iteratively reweighted TV regulariser for CS reconstruction. Spatially adaptive weights are computed towards a maximum a posteriori estimation of the image gradients. To exploit the nonlocal redundancy, effective nonlocal sparsity regularisation has also been introduced into the proposed objective function. Experimental results demonstrate that the proposed CS reconstruction method outperforms significantly existing TV-based CS reconstruction methods.

Compressive sensing via reweighted TV and nonlocal sparsity regularisation

Recent researches have shown that the adaptive directional lifting (ADL) can represent edges and textures in images effectively. This makes it possible to separate noise from image signal distinctly in image denoising. However, a key issue named orientation estimation for ADL becomes inefficient and error prone in the noised circumstance. The authors propose a robust adaptive directional lifting-based (RADL) wavelet transform for image denoising by constructing ADL in an anti-noise way. In our method, a simple model of pixel pattern classification is incorporated into orientation estimation module to strengthen the robustness of this algorithm. Moreover, instead of determining the transform strategy based on sub-blocks, RADL is performed on pixel-level to pursue better denoising results. Experimental results show that the proposed technique demonstrates both PSNR and visual quality improvement on images with rich textures.

Robust adaptive directional lifting wavelet transform for image denoising

This paper proposes an adaptive morphological dilation image coding with context weights prediction. The new dilation method is not to use fixed models, but to decide whether a coefficient needs to be dilated or not according to the coefficient's predicted significance degree. It includes two key dilation technologies: 1) controlling dilation process with context weights to reduce the output of insignificant coefficients, and 2) using variable-length group test coding with context weights to adjust the coding order and cost as few bits as possible to present the events with large probability. Moreover, we also propose a novel context weight strategy to predict coefficient's significance degree more accurately, which serves for two dilation technologies. Experimental results show that our proposed method outperforms the state of the art image coding algorithms available today.

Morphological dilation image coding with context weights prediction

In the latest High Efficiency Video Coding (HEVC) development, i.e., HEVC screen content coding extensions (HEVC-SCC), a hash-based inter-motion search/block matching scheme is adopted in the reference test model, which brings significant coding gains to code screen content. However, the hash table generation itself may take up to half the encoding time and is thus too complex for practical usage. In this paper, we propose a hierarchical hash design and the corresponding block matching scheme to significantly reduce the complexity of hash-based block matching. The hierarchical structure in the proposed scheme allows large block calculation to use the results of small blocks. Thus, we avoid redundant computation among blocks with different sizes, which greatly reduces complexity without compromising coding efficiency. The experimental results show that compared with the hash-based block matching scheme in the HEVC-SCC test model (SCM)-6.0, the proposed scheme reduces about 77% of hash processing time, which leads to 12% and 16% encoding time savings in random access (RA) and low-delay B coding structures. The proposed scheme has been adopted into the latest SCM. A parallel implementation of the proposed hash table generation on graphics processing unit (GPU) is also presented to show the high parallelism of the proposed scheme, which achieves more than 30 frames/s for 1080p sequences and 60 frames/s for 720p sequences. With the fast hash-based block matching integrated into x265 and the hash table generated on GPU, the encoder can achieve 11.8% and 14.0% coding gains on average for RA and low-delay P coding structures, respectively, for real-time encoding.

Guangming Shi

Papers

Nonlocal center-surround reconstruction-based bottom-up saliency estimation

Compressive sensing via reweighted TV and nonlocal sparsity regularisation

Robust adaptive directional lifting wavelet transform for image denoising

Morphological dilation image coding with context weights prediction

Fast Hash-Based Inter-Block Matching for Screen Content Coding