Realistic Virtual Reality is supported by 360° video, which provides viewers with an immersive watching experience. However, 360° video is bulky in size, while the transmission system has limited ability to provide bandwidth. As such, intelligent adaptive delivery solutions play a crucial role in enabling users to stream high-quality 360° video. In this research, we propose a novel approach for 360- degree video streaming over HTTP/2 that can provide consumers with a good watching experience (QoE) even in varying network circumstances and head-eye movements over time. The proposed method deploys the so-called BBAG algorithm (Buffer and Bandwidth Allocation Algorithm) using Scalable Video Coding to choose appropriate tile layers to resolve the trade-off between network and user adaptivity. With the support of HTTP/2’s stream termination capability, the delivery of late tile layers is terminated to handle abrupt interruptions. Our BBAG method is proven to improve QoE from 3.56% to 93.71% and the average viewport bitrate from 1.78% to 80.49% compared to state-of-the-art methods.

Face detection is a mandatory step in many computer vision applications, such as face recognition, emotion recognition, age detection, virtual makeup, and vital sign monitoring. Thanks to advancements in deep learning and the introduction of annotated large-scale datasets, numerous applications have been developed for human faces. Recently, other domains, such as animals and cartoon characters, have started gaining attention but still lag far behind human faces. The biggest challenge is the limited number of annotated face datasets in these domains. The manual labeling of large-scale datasets is tedious and requires substantial human labor. In this regard, we present an input-agnostic face detector to ease the annotation of various face datasets. We propose a simple but effective data-centric approach instead of building a specific neural network architecture. Specifically, we trained a face detection model, YOLO5Face, on human, animal, and cartoon face datasets. The experiments show that the model can achieve accurate results in all domains. In addition, the model achieved decent results for animals and cartoon characters different from the ones in the training set. This implies that the model can extract agnostic facial features. We have made the source code and pre-trained models publicly available at https://github.com/IS2AI/AnyFace to stimulate research in these fields.

Macroscopic superconductivity models have improved significantly over the last decades. Formulations and methods have been developed to faster solve problems involving high-temperature superconductors (HTS). Despite these developments, finite element analysis (FEA) of AC superconducting machines (SCMs) with HTS coils still remains time-consuming. The combined burden from the HTS models and the moving mesh makes the models complex and slow. To deal with this challenge, a new framework for FEA of AC superconducting SCMs with surface-mounted permanent magnet (PM) rotors and HTS armature windings is proposed in this paper. In this approach, the rotating rotor geometry is emulated with a stationary array of small PM segments excited with time-varying boundary sources. The major benefit of this approach is that the models can be realized without moving meshes, which increases the computation speed by more than one order of magnitude. In our dedicated case study of a complete SCM design, the speedup factors are 17.4x and 38.5x for the H-A and T-A formulations, respectively. Over a large parametric space of 18 design permutations, the highest relative error in calculated HTS loss was 4.12 percent. As a result, this work enables the designer to perform much more comprehensive performance studies of SCMs.

Abstract— This work-in-progress presents a compact toolkit for reproducing the training of T5-based multiinstruments and multi-track transcription models on multiple MIR tasks. We plan to release our code in early 2023 at https://github.com/mimbres/YourMT3.

This is a three-part series, where the numerical  discrete-time analysis and comparative analysis of 72 Ah,Lithium Ion Battery (LIB) is presented. The numerical analysis has been explained in detailed with two non-linear Kalman filter varients, namely Unscented Kalman Filter(UKF) and Extended Kalman Filter(EKF), both in combination with event based Kalman Filter algorithm. The discrete-time state equations have been developed for 8RC equivalent circuit model, to model any variant of LIB, irrespective of recommended application type. LIB has been modeled with 8 parallel RC branches with also taking into account various non-linear parameters,thermal(modeled with thermal mass, convection, ambient temperature variation),battery charge-disharge cycle degration effect and aging. The comparative results have been tabulated from 1 RC to 8 RC and compared with experimentally measured. 400 cells are arranged to give 72 Ah, where each cell has rating of 6.66 Wh. In the first part, the UKF algoritm along with event-based kalman filter has been explained with the derivation of equations.I m not completing research paper, the earlier version got lost in computer crash. One part is survived, uploaded it.I will not re-write or submit paper as left R&D.

A document by Taissir Elganimi . Click on the document to view its contents.

In this article, we analyzed the performance of several classes of quantum circuits under various errors. We executed NISQ circuits on both noisy simulator and real hardware backend publicly available from IBM Quantum. On simulation, we created noise models with both individual errors and combination of multiple errors to quantify how each type of error impact performance of quantum circuits. We utilized statistical distances such as total variation distance (TVD), Jensen-Shannon divergence (JSD), and Hellinger distance, to measure the deviation of noisy output from ideal output. Finally, we compare the noisy outputs from the simulator and outputs from the real hardware to showcase the gap between simulation and real device experiment. This gap points to the limitation of present noise models failure to exhibit the behavior of real hardware.

In this paper, a computationally efficient chance-constrained rollover-free motion planning method is presented. Specifically, the method is developed to plan motions for reconfigurable vehicles with the knowledge of a 3-D terrain model that has limited accuracy. The overall motion planning problem is formulated as a nonlinear optimal control problem (NOCP) that employs a constraint in the form of a bound on the probability of rollover under terrain-induced vehicle orientation uncertainty. To increase the computational efficiency of the NOCP, a geometric interpretation of the chance constraint is derived based on the characteristics of SO(3), the 3-D rotation group. Monte Carlo simulations are provided to demonstrate the usefulness of the geometric interpretation through comparisons with other methods. Experimental data gathered from driving a mobile robot through real forests are also used to validate the proposed model. Finally, path and trajectory generation results obtained with the proposed planning method for a feller-buncher machine traversing through uncertain 3-D terrain are presented to showcase the method’s overall performance and efficiency.

This research paper explores the concept of edge computing and the advancements in this technology. Edge computing is a distributed computing paradigm that brings computation and data storage closer to the devices that generate and collect data in order to reduce latency and improve performance. It has been identified as a powerful tool for providing low-latency access to data and applications and is becoming increasingly popular in the IoT and 5G era. This paper focuses on the various advancements in edge computing technology, such as the emergence of edge cloud platforms and the development of edge analytics and machine learning solutions. It also discusses various applications of edge com- puting, such as autonomous driving, augmented/virtual reality, and the Internet of Things, and how they are being used to improve efficiency and reduce latency. Finally, the paper analyses the various challenges and opportunities associated with edge computing, including security, scalability and privacy concerns.

A document by Siyue Li. Click on the document to view its contents.

The standard interpretation of Young’s double slit experiment is that the light behaves as waves before and after passing through the diaphragm of the double slit. Recently, the convex lens has been utilized to study the behavior of photons near the diaphragm. The experiments show that, within certain macroscopic distance, the pattern is the particle pattern, i.e., the non-interference pattern, and that the non-interference pattern gradually evolve to the interference pattern. blob:https://www.techrxiv.org/35fd3146-204c-4a6b-9148-f0f77f9af13d blob:https://www.techrxiv.org/44bec0b0-0f2c-46b6-addb-07e5678b17cd blob:https://www.techrxiv.org/c362971d-4441-4f4c-b275-8ef1202257fc In this article, we show that, in the double slit experiments, the particle pattern (non-interference pattern) and interference pattern emerge on the screen simultaneously. It is the challenge to interpret this phenomenon.

This paper aims to suggest method for sustaining vacuum in particle accelerator while in an variety of circumstances.

360° images and videos have become an economic and popular way to provide VR experiences using real-world content. However, the manipulation of the stereo panoramic content remains less explored. In this paper, we focus on the 360° image composition problem, and develop a solution that can take an object from a stereo image pair and insert it at a given 3D position in a target stereo panorama, with well-preserved geometry information. Our method uses recovered 3D point clouds to guide the composited image generation. More specifically, we observe that using only a one-off operation to insert objects into equirectangular images will never produce satisfactory depth perception and generate ghost artifacts when users are watching the result from different view directions. Therefore, we propose a novel view-dependent projection method that segments the object in 3D spherical space with the stereo camera pair facing in that direction. A deep depth densification network is proposed to generate depth guidance for the stereo image generation of each view segment according to the desired position and pose of the inserted object. We finally merge the synthesized view segments and blend the objects into the target stereo 360° scene. A user study demonstrates that our method can provide good depth perception and removes ghost artifacts. The view-dependent solution is a potential paradigm for other content manipulation methods for 360° images and videos.

Data transfer infrastructures composed of Data Transfer Nodes (DTN) are critical to supporting distributed computing and storage capabilities for clouds, data repositories, and complexes of supercomputers and instruments. The infrastructure’s throughput profile, estimated as a function of the connection round trip time using Machine Learning (ML) methods, is an indicator of its operational state, and has been utilized for monitoring, diagnosis and optimization purposes. We show that the inherent statistical variations and precision of the throughput profiles estimated by ML methods can be exploited for unauthorized use of DTNs’ computing and network capacity. We present a game theoretic formulation that captures the cost-benefit trade-offs between an attacker that attempts to hide under the profile’s statistical variations and a provider that counters by using additional measurements at an added cost. The Nash equilibrium conditions of this game provide qualitative insights and bounds for the success probabilities of the attacker and provider, based on the generalization equation of the ML-estimate. We present experimental results that illustrate scenarios wherein a significant portion of DTN computing capacity is compromised without being detected by an attacker that exploits the ML estimate properties.

A broadband noise measurement system is newly developed and demonstrated at temperatures between 3 K and 300 K. Using the system, wideband noise spectroscopy (WBNS) from 20 kHz to 500 MHz is carried out for the first time, revealing that shot noise is the dominant white noise down to 3 K. The paper also suggests, by means of WBNS, the possibility of extracting the baseline noise characteristics, which do not include the noise component that varies a great deal from device to device.

Driving requires continuous decision making from a driver taking into account all available relevant information. Automating driving tasks also automates the related decisions. However, humans are very good at dealing with bad quality, fuzzy, informal and incomplete information, whereas machines generally require solid quality information in a formalized format. Therefore, the development of automated driving functions relies on the availability of machine-usable information. A digital twin contains quality controlled information collected and augmented from different sources, ready to be supplied to such an automated driving function. An information model that describes all conceivably relevant information is necessary. To this end, a list of requirements that such an information model should meet is proposed and each requirement is argued for. Based on the anticipated services and applications that such a system should support, a collection of requirements for system architecture is derived. Information modeling is performed for selected relevant information groups. A system architecture has been proposed and validated with three different implementations, addressing several different applications to support decisions at a highway tunnel construction site in Austria and throughout the Test Bed Lower Saxony in Germany.

Forest inventory forms the foundation of forest management. Remote sensing (RS) is an efficient means of measuring forest parameters at scale. Remotely sensed species classification can be used to estimate species abundances, distributions, and to better approximate metrics such as above ground biomass.  State of the art methods of RS species classification rely on deep learning models such as convolutional neural networks (CNN). These models have 2 major drawbacks: they require large samples of each species to classify well and they lack explainablity. Therefore, rare species are poorly classified causing poor approximations of their associated parameters. We show that the classification of rare species can be improved by as much as 8 F1-points using a neuro-symbolic (NS) approach that combines CNNs with a NS framework. The framework allows for the incorporation of domain knowledge into the model through the use of mathematically represented rules, improving model explainability.

The determinization of fuzzy automata is a well-studied problem in theoretical computer science celebrated for its practical applications. Indeed, in the fields of fuzzy discrete event systems, fault diagnosis, clinical monitoring, decision-making systems, and model checking, when a suitable model of a fuzzy automaton is employed, it is desirable to find its language-equivalent deterministic version because of its computational efficiency. Although many methods have been developed to convert a fuzzy automaton to its language equivalent fuzzy deterministic finite automaton (FDfA), they can be applied only for fuzzy automata defined over specific underlying sets of truth values. For example, recently developed determinization methods employ the concept of maximal factorization, which can be defined only on non-locally finite lattices or the Boolean lattice. In addition, not all such determinization methods result in a minimal FDfA. On the other hand, even such determinization methods have been developed for fuzzy automata over specific underlying structures, these methods cannot be generalized for fuzzy automata over locally finite lattices. This article focuses on filling this gap and develops a novel method for computing a minimal FDfA for a fuzzy automaton defined over a locally finite and divisible residuated lattice. Our method uses the concept of a reduction graph that helps us achieve a minimal FDfA.