Abstract: User expectation is one of the main factors that drives the user satisfaction for video streaming service providers and online social media platforms.Depending on the context, users may have different expectations of the video quality.Measuring the Quality of Experience (QoE) by taking user expectations into account provide online social media platforms with increased efficiency and users with higher satisfaction.In this work, we explore the relation between video quality and acceptability&annoyance of video quality in online social media platforms context.Moreover we present the methodology to determine the metric thresholds for acceptability&annoyance of video quality.We compare the estimated thresholds with previous studies.

Abstract: Different subjective methodologies exist to collect data on human perception of distortions, from rating methodologies with single or double stimuli to ranking with pairwise comparisons.The Maximum Likelihood Difference Scaling (MLDS) method uses triplet/quadruplet-based comparisons as a ranking task.Participants compare intervals inside pairs of stimuli: (a,b) and (c,d).The task is to rank if they perceive greater differences between (a,b) or (c,d).From these comparisons' judgments, we can place the assessed stimuli on a perceptual scale (e.g., from low to high quality) with the help of a mathematical solver.However, one limitation is that the perceptual scales retrieved from stimuli of multiple contents are usually different.We previously offered a solution to measure the inter-content scale of multiple contents.In this work, we compare multiple rating and ranking methodologies.We examine obtained subjective quality scores regarding precision by analyzing discriminability in the scores, efficiency by comparing fixed experimental effort costs, and robustness of retrieve estimates to outliers and spammer behaviors.In this work, we put data quality, experimental cost, and resolving power into relation.We show how discriminability in the data impacts the resolving power of popular objective quality metrics.Our findings are that higher-performing metrics require higher-quality data to reveal their full potential.

Abstract: We introduce an efficient distributed sequence parallel approach for training transformer-based deep learning image segmentation models.The neural network models are comprised of a combination of a Vision Transformer encoder with a convolutional decoder to provide image segmentation mappings.The utility of the distributed sequence parallel approach is especially useful in cases where the tokenized embedding representation of image data are too large to fit into standard computing hardware memory.To demonstrate the performance and characteristics of our models trained in sequence parallel fashion compared to standard models, we evaluate our approach using a 3D MRI brain tumor segmentation dataset.We show that training with a sequence parallel approach can match standard sequential model training in terms of convergence.Furthermore, we show that our sequence parallel approach has the capability to support training of models that would not be possible on standard computing resources.

Abstract: A new algorithm for the detection of deepfakes in digital videos is presented. The I-frames were extracted in order to provide faster computation and analysis than approaches described in the literature. To identify the discriminating regions within individual video frames, the entire frame, background, face, eyes, nose, mouth, and face frame were analyzed separately. From the Discrete Cosine Transform (DCT), the β components were extracted from the AC coefficients and used as input to standard classifiers. Experimental results show that the eye and mouth regions are those most discriminative and able to determine the nature of the video under analysis.

Abstract: This paper presents a method for synthesizing 2D and 3D sensor data for various machine vision tasks.Depending on the task, different processing steps can be applied to a 3D model of an object.For object detection, segmentation and pose estimation, random object arrangements are generated automatically.In addition, objects can be virtually deformed in order to create realistic images of non-rigid objects.For automatic visual inspection, synthetic defects are introduced into the objects.Thus sensor-realistic datasets with typical object defects for quality control applications can be created, even in the absence of defective parts.The simulation of realistic images uses physically based rendering techniques.Material properties and different lighting situations are taken into account in the 3D models.The resulting tuples of 2D images and their ground truth annotations can be used to train a machine learning model, which is subsequently applied to real data.In order to minimize the reality gap, a random parameter set is selected for each image, resulting in images with high variety.Considering the use cases damage detection and object detection, it has been shown that a machine learning model trained only on synthetic data can also achieve very good results on real data.

Abstract: Starch plays a pivotal role in human society, serving as a vital component of our food sources and finding widespread applications in various industries.Microscopic imaging offers a straightforward, efficient, and precise approach to examine the distribution, morphology, and dimensions of starch granules.Quantitative analysis through the segmentation of starch granules from the background aids researchers in exploring their physicochemical properties.This article presents a novel approach utilizing a modified U-Net model in deep learning to achieve the segmentation of starch granule microscope images with remarkable accuracy.The method yields impressive results, with mean values for several evaluation metrics including JS, Dice, Accuracy, Precision, Sensitivity and Specificity-reaching 89.

Abstract: In the field of polymers, 2D images are often used to discern information about the microstructure of bulk polymer materials.For brush particle assembly structures, this work evaluates microstructure information retrieved from different material characterization techniques for thin film (i.e., electron imaging of brush particle monolayers) and bulk materials (small angle X-ray scattering), respectively.The effect of confinement of polymer chains into thin (2D) films on the conformation of tethered chains is discussed and used to rationalize systematic discrepancies between characteristic nanoparticle spacings in thin films and bulk materials.An approach to rationalize bulk material properties based on thin film measurements is presented.

Abstract: The integration of trench vertical transfer gates in an indirect time of flight pixel has been studied through TCAD & optical simulations.A small fast photo-gate pixel surpassing state of the art performances has been designed and optimized thanks to these advanced multiphysics simulations.Quantum efficiency of 40% is obtained and demodulation contrast of 89% at 200MHz is achieved while transfer gates operate at 1.0V biasing.

Abstract: Video DeepFakes are fake media created with Deep Learning (DL) that manipulate a person’s expression or identity. Most current DeepFake detection methods analyze each frame independently, ignoring inconsistencies and unnatural movements between frames. Some newer methods employ optical flow models to capture this temporal aspect, but they are computationally expensive. In contrast, we propose using the related but often ignored Motion Vectors (MVs) and Information Masks (IMs) from the H.264 video codec, to detect temporal inconsistencies in DeepFakes. Our experiments show that this approach is effective and has minimal computational costs, compared with per-frame RGB-only methods. This could lead to new, real-time temporally-aware DeepFake detection methods for video calls and streaming.

Abstract: In today's cybersecurity landscape, protecting information systems is crucial due to the rising threat of cyber-attacks.This research focuses on vulnerability management using open-source tools for domain and subdomain enumeration, vulnerability scanning, and remediation.Open-source software offers costeffective and collaborative security solutions.Domain and subdomain enumeration tools play a vital role in mapping an organization's attack surface, providing insight into security posture.The analysis of vulnerability scanning tools highlights their effectiveness in identifying critical flaws in web applications and databases.Vulnerability remediation through patching, hardening, and exposure management processes closes security gaps.The research provides an empirical insight into using open-source tools for vulnerability management, listing their benefits and limitations empowering organizations to enhance their security posture.Recommendations for integrating these tools into existing security frameworks help combat cyber threats and protect valuable assets.1. To identify a variety open-source tools for asset discovery, to map out the attack surface of target organization by enumerating domains, subdomains and ASN's through passive means, their scope and limitations and a comparative analysis the result of their enumerations.2. To identify and test a range of open-source vulnerability scanners and conduct a comparative analysis of the scanners by testing them against test website http://testphp.vulnweb.com/,http://php.testsparker.com/and comparing the types of issues found and how we can use them to identify vulnerabilities in our assets.3.

Abstract: ACES [1] is a standardized color management system widely used in the film and visual effects industry to ensure consistent and accurate color reproduction throughout the production pipeline.Integrating ACES into game engines like Unreal Engine [2] could have significant benefits, especially for game developers who want to achieve high-quality, consistent color representation across different platforms and displays. Game developers can achieve heightened visual fidelity by leveraging ACES in Unreal Engine 5, especially concerning wide color gamuts and high dynamic range (HDR) content. The standardized color management system allows cross-platform development, guaranteeing consistent color reproduction on various devices and display technologies. Moreover, Unreal Engine 5's support for ACES facilitates seamless collaboration with other creative industries that utilize this industry-standard color pipeline.However, implementing ACES in a real-time engine presents unique challenges regarding performance optimization and ensuring compatibility with other game engines.Artists and developers may need to adapt their workflows to accommodate ACES color transforms, impacting the art pipeline and usergenerated content.This paper uses ACES to investigate color input and output consistency to and from Epic Games Unreal 5 regarding Wide Color Gamut and High Dynamic Range imagery.

Abstract: This paper proposes a novel aggregation method using the Kumaraswamy distribution to analyze partial metric values, particularly in the evaluation of video quality.Through a weighted mean aggregation procedure, we unravel the underlying effects on the data.The three experiments analyzed in this paper demonstrates the method's efficacy regardless of the time aggregation, ranging from days, minutes, and frames.This approach, grounded in the Kumaraswamy distribution, offers a robust analytical tool to understand how individual metric values amalgamate, affecting overall user perceptions and experience.

Abstract: Event-based vision Sensors (EVS) utilize smart pixels capable of detecting whether relative illumination changes exceed a predefined temporal contrast threshold on a pixel level.As EVS asynchronously read these events, they provide low-latency and high-temporal resolution suitable for complementing conventional CMOS Image Sensors (CIS).Emerging hybrid CIS+EVS sensors fuse the high spatial resolution intensity frames with low latency event information to enhance applications such as deblur or video-frame interpolation (VFI) for slow-motion video capture.This paper employs an edge sharpness-based metric-Blurred Edge Width (BEW) to benchmark EVS-assisted slow-motion capture against CIS-only solutions.The EVS-assisted VFI interpolates a CIS video steam with a framerate of 120 fps by 64x, yielding an interpolated framerate of 7680 fps.We observed that the added information from EVS dramatically outperforms a 120 fps CIS-only VFI solution.Furthermore, the hybrid EVS+CIS-based VFI achieves comparable performance as high-speed CIS-only solutions that capture frames directly at 480 fps or 1920 fps and incorporate additional CIS-only VFI.These, however, do so at significantly lower data rates.In our study, factors ∼ 2.6 and ∼ 10.5 were observed.

Abstract: This paper proposes a pixel-wise parameter estimation framework for Event-based Vision Sensor (EVS) characterization.Using an ordinary differential equation (ODE) based pixel latency model and an autoregressive Monte-Carlo noise model, we first identify the representative parameters of EVS.The parameter estimation is then formulated as an optimization problem to minimize the measurement-prediction error for both pixel latency and event firing probability.Finally, the effectiveness and accuracy of the proposed framework are verified by comparison of synthetic and measured event response latency as well as firing probability as function of temporal contrast (so-called S-curves).

Abstract: Aberrations in the optical system of a light field (LF) display can degrade its quality and affect the focusing effects in the retinal image, formed by the superposition of multiperspective LF views.To address this problem, we propose a method for calibrating and correcting aberrated LF displays.We employ an LF display optical model to subsequently derive the retinal image formation with a given LF input.This enables us to efficiently measure actual viewpoint locations and the deformation of each perspective image, by capturing focal-stack images during the calibration process.We then use the calibration coefficients to pre-warp the input images so that the aberrations are compensated.We demonstrate the effectiveness of our method on a simulated model of an aberrated near-eye LF display and show that it improves the display's optical quality and the accuracy of the focusing effects.

Abstract: The fast growth of cloud computing technology has led to immense development in the public and private sectors.Cloud computing provides a high level of virtualization, massive scalability, multitenancy, and elasticity.This has enabled organizations, academia, government departments, and the public to advance with this technology.However, they cannot assuredly place their information in the cloud due to many security threats.Cloud security plays a vital role in establishing confidence between the cloud service providers, consumers, and multi-users to maintain the security levels of their data.Moreover, in the scope of cloud computing, the importance of security testing must be considered.Security testing involves evaluating the cloud infrastructure and applications for vulnerabilities, ensuring that sensitive data remains protected.This paper focused on the challenges, tools, techniques, and methodologies for cloud security testing.Furthermore, the paper introduces the tools offered by three significant CSPs for cloud security testing and the most critical cloud vulnerabilities.It explains some published vulnerabilities around these three major CSPs.Between these three significant CSPs, we focused on Azure offerings for securing their clouds and some known tools for security testing in the cloud.Lastly, we introduced and explained the most essential API vulnerabilities according to OWASP and a suggested way to mitigate them. Literature ReviewIn the realm of cloud security testing, significant advancements have been made, particularly in the development of automated testing systems.Tao, Lin, and Lu (2015) designed a cloud platform-based automated testing system specifically for the mobile internet environment.This system leverages virtualization and automation technology to integrate mobile terminals into the cloud platform, offering a novel service known as Testing as a Service (TaaS).The system's ability to flexibly

Abstract: This paper presents a prototype that aims to make education accessible to all.The chosen learning topic focuses on youth language to overcome language barriers and create better access to society for those who are not fluent in German.The results are based on a systematic literature review and the development of a prototype tested using the BITV tests.The key findings are that accessibility should be a strategic goal set and repeatedly promoted by senior management.This way, it does not become less of a priority, which has often proven to be a weakness because the accessible elements of software are not necessarily needed for its core functions.With the insight of using an agile participation model, accessible applications can be completed in a shorter average time in the future.The project also demonstrates the importance of accessibility awareness in software development.

Abstract: Assuming that Alice commits to an embedding method and the Warden to a detector, we study how much information Alice can communicate at a constant level of statistical detectability over potentially infinitely many uses of the stego channel.When Alice is allowed to allocate her payload across multiple cover objects, we find that certain payload allocation strategies that are informed by a steganography detector exhibit super-square root secure payload (scaling exponent 0.85) for at least tens of thousands of uses of the stego channel.We analyze our experiments with a source model of soft outputs of the detector across images and show how the model determines the scaling of the secure payload. Response curveWe use C to denote the maximum embedding capacity of a cover image X ∈ X .For a ternary embedding scheme in the spatial domain, C ≤ log 2 3 bits per pixel (bpp).Since most steganographic schemes avoid making changes to saturated pixels, the capacity can be strictly smaller than log 2 3.

Abstract: Dense 3D reconstruction has many applications in automated driving including automated annotation validation, multimodal data augmentation, providing ground truth annotations for systems lacking LiDAR, as well as enhancing auto-labeling accuracy.LiDAR provides highly accurate but sparse depth, whereas camera images enable estimation of dense depth but noisy particularly at long ranges.In this paper, we harness the strengths of both sensors and propose a multimodal 3D scene reconstruction using a framework combining neural implicit surfaces and radiance fields.In particular, our method estimates dense and accurate 3D structures and creates an implicit map representation based on signed distance fields, which can be further rendered into RGB images, and depth maps.A mesh can be extracted from the learned signed distance field and culled based on occlusion.Dynamic objects are efficiently filtered on the fly during sampling using 3D object detection models.We demonstrate qualitative and quantitative results on challenging automotive scenes.

Abstract: Domain Adaptation (DA) techniques aim to overcome the domain shift between a source domain used for training and a target domain used for testing.In recent years, vision transformers have emerged as a preferred alternative to Convolutional Neural Networks (CNNs) for various computer vision tasks.When used as backbones for DA, these attention-based architectures have been found to be more powerful than standard ResNet backbones.However, vision transformers require a larger computational overhead due to their model size.In this paper, we demonstrate the superiority of attention-based architectures for domain generalization and source-free unsupervised domain adaptation.We further improve the performance of ResNet-based unsupervised DA models using knowledge distillation from a larger teacher model to the student ResNet model.We explore the efficacy of two frameworks and answer the question: is it better to distill and then adapt or to adapt and then distill?Our experiments on two popular datasets show that adapt-to-distill is the preferred approach.

Abstract: Critical infrastructure is the backbone of modern societies, and protecting this infrastructure is essential to ensure the stability of societies and economies.The electricity sector is one of the most critical infrastructures, and any disruption can have significant consequences.The threat landscape in this sector is constantly evolving.With the increasing sophistication of cyberattacks and other threats, it has become essential to use innovative technologies to identify and mitigate them.Open Source Intelligence (OSINT) technologies have emerged and offer valuable tools for identifying and mitigating these threats.This article presents an in-depth overview of OSINT technologies and their applications in the protection of critical infrastructure, with an emphasis on the electricity sector.It discusses the vulnerabilities of the electricity sector, the types of OSINT technologies, and the benefits they provide.Case studies of successful applications of OSINT technologies in the electricity sector are presented to illustrate their effectiveness.This article also examines organizations' challenges in implementing OSINT technologies, including technological, legal, and financial challenges.Finally, the article concludes by offering recommendations for successfully implementing OSINT technologies to protect critical infrastructure, particularly in the electricity sector.The insights offered in this article will be helpful for policymakers, security professionals, and anyone interested in protecting critical infrastructure.