Graphs are fundamental mathematical structures used in various fields to represent data, signals, and processes. In this paper, we propose a novel framework for learning/estimating graphs from data. The proposed includes (i) formulation of graph learning problems, (ii) their probabilistic interpretations, (iii) associated algorithms. Specifically, problems posed as the estimation Laplacian matrices some observed data under given structural constraints (e.g., connectivity sparsity level)....

OpenFlow is a programmable network protocol and associated hardware designed to effectively manage direct traffic by decoupling control forwarding layers of routing. This paper presents an analytical framework for optimization decisions at the layer enable dynamic Quality Service (QoS) over networks discusses application this QoS-enabled streaming scalable encoded videos with two QoS levels. We pose solve routing as constrained shortest path problem, where we treat base video level-1 flow,...

This paper presents novel QoS extensions to distributed control plane architectures for multimedia delivery over large-scale, multi-operator Software Defined Networks (SDNs). We foresee that large-scale SDNs shall be managed by a consisting of multiple controllers, where each controller performs optimal routing within its domain and shares summarized (aggregated) information with other controllers enable inter-domain reduced problem dimensionality. To this effect, proposes (i) topology...

In the past decade, development of transform coding techniques has achieved significant progress and several advanced tools have been adopted in new generation Versatile Video Coding (VVC) standard. this paper, a brief history during VVC standardization is presented, standard are described detail together with their initial design, incremental improvements implementation aspects. To improve efficiency, four introduced VVC, which namely Multiple Transform Selection (MTS), Low-Frequency...

OpenFlow is a clean-slate Future Internet architecture that decouples control and forwarding layers of routing, which has recently started being deployed throughout the world for research purposes. This paper presents an optimization framework controller in order to provide QoS support scalable video streaming over network. We pose solve two problems, where we route base layer SVC encoded as lossless-QoS flow, while enhancement can be routed either lossy-QoS flow or best effort respectively....

This paper introduces a novel spectral anomaly detection method by developing graph-based filtering framework. In particular, we consider the problem of unsupervised data over wireless sensor networks (WSNs) where measurements are represented as signals on graph. our framework, graphs chosen to capture useful proximity information about measured data. The associated filters then employed project graph normal and subspaces, resulting projections used in anomalies. proposed approach has two...

This paper introduces a novel graph signal processing framework for building graph-based models from classes of filtered signals. In our framework, modeling is formulated as system identification problem, where the goal to learn weighted (a Laplacian matrix) and filter function matrices). order solve proposed an algorithm developed jointly identify (GBF) multiple signal/data observations. Our valid under assumption that GBFs are one-to-one functions. The approach can be applied diffusion...

Recent papers have formulated the problem of learning graphs from data as an inverse covariance estimation with graph Laplacian constraints. While such problems are convex, existing methods cannot guarantee that solutions will specific topology properties (e.g., being $k$-partite), which desirable for some applications. In fact, a given properties, e.g., finding $k$-partite best matches data, is in general non-convex. this paper, we develop novel theoretical results provide performance...

The Karhunen-Loeve transform (KLT) is known to be optimal for decorrelating stationary Gaussian processes, and it provides effective coding of images. Although the KLT allows efficient representations such signals, itself completely data-driven computationally complex. This paper proposes a new class transforms called graph template (GTTs) that approximate by exploiting priori information about signals represented graph-template. In order construct GTT (i) design matrix leading defined, then...

In traditional image and video coding schemes, separable transforms are typically employed due to their low-complexity implementations. However, the compression efficiency of is limited for most natural image/video blocks which generally have arbitrarily directed edge texture patterns. It well known that non-separable can achieve better directional patterns, yet they computationally complex, especially larger block sizes. order higher transform gains with relatively implementations, in this...

In video coding, motion compensation is an essential tool to obtain residual block signals whose transform coefficients are encoded. This paper proposes novel graph-based transforms (GBTs) for coding inter-predicted signals. Our contribution twofold: (i) We develop edge adaptive GBTs (EA-GBTs) derived from graphs estimated blocks, and (ii) we design template (TA-GBTs) by introducing simplified graph templates generating different set of with low signaling overhead. experimental results show...

This paper introduces a novel class of transforms, called graph-based separable transforms (GBSTs), based on two line graphs with optimized weights. For the optimal GBST construction, we formulate graph learning problem to design separate using row-wise and column-wise residual block statistics, respectively. We also analyze optimality resulting for both intra inter predicted models. Moreover, show that DCT ADST (DST-7) are special cases GBSTs. Our experimental results demonstrate proposed...

This paper proposes a new Quality of Service (QoS) optimized routing architecture for video streaming over large-scale multi-domain OpenFlow networks managed by distributed control plane, where each controller performs optimal within its domain and shares summarized intra-domain data with other controllers to reduce problem dimensionality calculating inter-domain routing. We apply the proposed scalable (layered) videos, base layer routes are dynamically fulfill required QoS level, while...

Graphs are fundamental mathematical structures used in various fields to represent data, signals and processes. In this paper, we propose a novel framework for learning/estimating graphs from data. The proposed includes (i) formulation of graph learning problems, (ii) their probabilistic interpretations (iii) associated algorithms. Specifically, problems posed as estimation Laplacian matrices some observed data under given structural constraints (e.g., connectivity sparsity level). From...

Most of the existing deep learning based end-to-end image/video coding (DLEC) architectures are designed for non-subsampled RGB color format. However, in order to achieve a superior performance, many state-of-the-art block-based compression standards such as High Efficiency Video Coding (HEVC/H.265) and Versatile (VVC/H.266) primarily YUV 4:2:0 format, where U V components subsampled by considering human visual system. This paper investigates various DLEC designs support format comparing...

The graph Fourier transform (GFT) - adaptive to the signal structures of local pixel blocks has recently been shown be a good alternative fixed transforms, e.g., Discrete Cosine Transform (DCT), for image coding. However, majority proposed GFTs assume an underlying 4-connected structure with vertical and horizontal edges only. In this paper, we propose design methodology select more general sparse edge weights, on which are defined block-based Specifically, first cluster via Lloyd-Max...

This demo abstract describes an initial design of a new adaptive video streaming protocol for device-to-device WiFi-based mobile platforms and its software implementation. For the demonstration, two servers users will be deployed verifying that our implementation works with desirable user experience.

Graphs are fundamental mathematical structures used in various fields to represent data, signals and processes. This paper proposes a novel framework for learning graphs from data. The proposed (i) poses the graph problem as estimation of generalized Laplacian matrices (ii) develops an efficient algorithm. Under specific statistical assumptions, formulation leads modeling attractive Gaussian Markov random fields. Our experimental results show that algorithm outperforms sparse inverse...

In existing video coding standards such as H.264/AVC and HEVC, the intra prediction is typically derived using fixed, symmetric filters along direction, e.g., in planar mode, top-right bottom-left samples are predicted filters. However, case of asymmetric availability neighboring reference samples, performance designed HEVC may not be optimal. To further refine achieve higher accuracy this paper proposes low-complexity refinements over prediction, which applied on frequently used planar, DC,...

In many state-of-the-art compression systems, signal transformation is an integral part of the encoding and decoding process, where transforms provide compact representations for signals interest. This paper introduces a class called graph-based (GBTs) video compression, proposes two different techniques to design GBTs. first technique, we formulate optimization problem learn graphs from data solutions optimal separable nonseparable GBT designs, GL-GBTs. The optimality proposed GL-GBTs also...

For the last few decades, application of signal-adaptive transform coding to video compression has been stymied by large computational complexity matrix-based solutions. In this paper, we propose a novel parametric approach greatly reduce without degrading performance. our approach, instead following conventional technique identifying full matrices that yield best efficiency, look for parameters defining new class transforms, called HyGTs, which have low implementations are easy parallelize....

Spectrum sensing is an essential functionality of cognitive radio wireless networks (CRWNs) that enables detecting unused frequency sub-bands for dynamic spectrum access. This paper proposes a compressed framework by (i) constructing sparsity basis in wavelet domain helps at sub-Nyquist rates and (ii) applying wavelet-based singularity detector on the reconstructed signal to identify available with low complexity. In particular, sensing, optimized Haar employed sparsely represent piecewise...

This paper introduces row-column transforms (RCTs) which are 2D non-separable defined with the aid of a set 1-D linear and basis ordering permutation. We propose novel method for design that approximate desired complex (such as KLTs, SOTs, etc.) so most performance approximated is retained at significantly reduced complexity. Given block transform interest, our designs an RCT by (i) optimizing applied to rows columns signal (ii) finding best coefficient Our experimental results show...