In this paper we compare the performance of two state-of-the-art competing codecs, AV1 and HEVC, in context adaptive streaming. We specifically consider a Dynamic Optimizer (DO) methodology that is content-aware selects resolution video sequence after constructing convex hull Rate-Quality curves all considered resolutions. start with an objective evaluation Optimizer, based on both PSNR VMAF quality metrics. The Rate-VMAF show average 6.3% BD-Rate gain over while Rate-PSNR loss 1.8%. then...

A challenge that many video providers face is the heterogeneity of networks and display devices for streaming, as well dealing with a wide variety content different encoding performance. In past, fixed bit rate ladder solution based on "fitting all" approach has been employed. However, such content-tailored highly demanding; computational financial cost constructing convex hull per by at all resolutions quantization levels huge. this paper, we propose content-gnostic exploits machine...

Accurate prediction of patient prognosis can be especially useful for the selection best treatment protocols. Machine Learning serve this purpose by making predictions based upon generalizable clinical patterns embedded within learning datasets. We designed a study to support feature 2-year prognostic period and compared performance several algorithms accurate estimation in advanced-stage high grade serous ovarian cancer (HGSOC) patients.The was formulated as binary classification problem....

One of the challenges faced by many video providers is heterogeneity network specifications, user requirements, and content compression performance. The universal solution a fixed bitrate ladder inadequate in ensuring high quality experience without re-buffering or introducing annoying artifacts. However, content-tailored solution, based on extensively encoding across all resolutions over wide range highly expensive terms computational, financial, energy costs. Inspired this, we propose an...

The impact of Artificial Intelligence (AI) is transforming various aspects urban life, including, governance, policy and planning, healthcare, sustainability, economics, entrepreneurship, etc. Although AI immense potential for positively impacting living, its success depends on overcoming significant challenges, particularly in telecommunications infrastructure. Smart city applications, such as, federated learning, Internet Things (IoT), online financial services, require reliable Quality...

In this paper, a novel spatial resolution adaptation approach for video compression is proposed. Its ability to dynamically apply downsampling frames exhibiting low detail delivers improved rate distortion performance, together with reduction in computational complexity of the encoding process. This method based on an experimental investigation dependence between QP threshold, which determines when encode lower frames, and obtained after downsampling/upsampling. The proposed integrated High...

In HTTP Adaptive Streaming, video content is conventionally encoded by adapting its spatial resolution and quantization level to best match the prevailing network state display characteristics. It well known that traditional solution, of using a fixed bitrate ladder, does not result in highest quality experience for user. Hence, this paper, we introduce content-driven approach estimating based on spatio-temporal features extracted from uncompressed content. The method implements...

The environmental impact of video streaming services has been discussed as part the strategies towards sustainable information and communication technologies. A first step that is energy profiling assessment consumption existing This paper presents a comprehensive study power measurement techniques for encoding decoding comparing use hardware software meters. An experimental methodology to ensure reliability measurements introduced. Key findings demonstrate high correlation based case two...

This work addresses the problem of predicting compression efficiency a video codec solely from features extracted uncompressed content. Towards this goal, we have used database videos homogeneous texture and both spatial frequency domain features. The are encoded using High Efficiency Video Coding (HEVC) reference at different quantization scales their Rate-Distortion (RD) curves modelled linear regression. Using fitted parameters RD model, Support Vector Regression Model (SVRM) is trained...

The adoption of video conferencing and communication services, accelerated by COVID-19, has driven a rapid increase in data traffic. demand for higher resolutions quality, the need immersive formats, newest, more complex codecs energy consumption centers display devices. In this paper, we explore compare across optimized state-of-the-art codecs, SVT-AV1, VVenC/VVdeC, VP9, x.265. Furthermore, align usage with various objective quality metrics compression performance set sequences different...

In this paper, an extensive study of different video texture properties based on encoding statistics extracted from the HEVC HM reference software is presented. Mode selection, partitioning, motion vectors and bitrate allocation are among obtained encoder. For study, a new dataset homogeneous static dynamic textures, HomTex, proposed. A comprehensive investigation results reveals significant variability coding within suggesting that category should be further split into two relevant...

Highly textured video content is challenging to compress since the bit-rate quality trade-off high and complex perceptual masking influences performance. Test datasets that cover a wide range of texture types are thus important for codec evaluation, but few exist. In order study properties texture, this paper introduces Synthetic Texture dataset (BVI-SynTex) was generated using Computer-Generated Imagery (CGI) environment. It contains 196 sequences clustered in three different offers...

During recent years, the standardisation committees on video compression and broadcast formats have worked extending practical frame rates up to 120 frames per second. Generally, increased been shown improve immersion, but at cost of higher bit rates. Taking into consideration that benefits high are content dependent, a decision mechanism recommends appropriate rate for specific would provide prior transmission. Furthermore, this must take account perceived quality. The proposed method...

This paper investigates quality-driven cross-layer optimization for resource allocation in direct sequence code division multiple access wireless visual sensor networks. We consider a single-hop network topology, where each transmits directly to centralized control unit (CCU) that manages the available resources. Our aim is enable CCU jointly allocate transmission power and source-channel coding rates node, under four different criteria take into consideration varying motion characteristics...

Encoding spatio-temporally varying textures is challenging for standardised video encoders, with significantly more bits required textured blocks compared to non-textured blocks. It therefore beneficial understand in terms of both their spatio-temporal characteristics and encoding statistics order optimize coding modes performance. To this end, we examine the classification texture based on encoder For purpose, employ features follow a two-step feature selection process by employing...

We propose a novel approach for the optimized network resource management of Direct Sequence Code Division Multiple Access (DS-CDMA) visual sensor network. The sensors monitor different scenes varying motion levels, thus resources need to be allocated each sensor. For recorded scene, our considers its individual content-related parameters, in contrast with previous methods that group according amount present scene and assign same transmission parameters all members group. Cross-layer...

The video technology scenery has been very vivid over the past years, with novel coding technologies introduced that promise improved compression performance state-of-the-art technologies. Despite fact a lot of datasets are available, representative content wide parameter space along subjective evaluations variations encoded from an unpartial end is required. In response to this requirement, paper features dataset, BVI-CC. Three codecs were deployed create sequences: High Efficiency Video...

We propose a novel priority-based approach that enables the optimal control of transmission power and use available network resources multihop Direct Sequence Code Division Multiple Access (DS-CDMA) Wireless Visual Sensor Network (WVSN). TheWVSN nodes can either monitor different scenes (source nodes) or retransmit videos other (relay nodes). Moreover, in real environments source may be dissimilar importance. Hence higher end-to-end quality is demanded for those are assigned priority....

VMAF is a machine learning based video quality assessment method, originally designed for streaming applications, which combines multiple metrics and features through SVM regression. It offers higher correlation with subjective opinions compared to many conventional methods. In this paper we propose enhancements the integration of new alternative (selected from diverse pool) alongside model combination. The proposed combination approach enables training on databases varying content...