As Artificial Intelligence (AI) is becoming ubiquitous in many applications, serverless computing also emerging as a building block for developing cloud-based AI services. Serverless has received much interest because of its simplicity, scalability, and resource efficiency. However, due to the trade-off with efficiency, suffers from cold start problem, that is, latency between request arrival function execution. The problem significantly influences overall response time workflow consists...

Vision Transformers (ViTs) have achieved remarkable success in various computer vision tasks. However, ViTs a huge computational cost due to their inherent reliance on multi-head self-attention (MHSA), prompting efforts accelerate for practical applications. To this end, recent works aim reduce the number of tokens, mainly focusing how effectively prune or merge them. Nevertheless, since ViT tokens are generated from non-overlapping grid patches, they usually do not convey sufficient...

We report on efficient generation of ultra-broadband terahertz (THz) waves via optical rectification in a novel nonlinear organic crystal with acentric core structure, i.e. 2-(4-hydroxystyryl)-1-methylquinolinium 4-methylbenzenesulfonate (OHQ-T), which possesses an ideal molecular structure leading to maximized response for near-infrared-pumped THz wave generation. By systematic studies wavelength-dependent phase-matching conditions OHQ-T crystals different thicknesses we are able generate...

Solving partial differential equations (PDEs) by learning the solution operators has emerged as an attractive alternative to traditional numerical methods. However, implementing such architectures presents two main challenges: flexibility in handling irregular and arbitrary input output formats scalability large discretizations. Most existing are limited their desired structure or infeasible scale inputs outputs. To address these issues, we introduce attention-based model called inducing...

We report a new crystal engineering technique for the thickness-controlled growth of large-area organic electro-optic crystals in confined geometry optimal THz wave photonics. investigate relationship between material properties and generation characteristics to determine applications. Two different state-of-the-art nonlinear optical are chosen as model crystals, ionic quinolinium-based non-ionic polyene-based crystals. Both kinds grown exhibit good quality, large area sub-millimeter-scale...

New phenolic quinolinium crystals with very large macroscopic optical nonlinearity and good crystal-growth ability exhibit an out-of-plane polar axis, which is unusual considering other state-of-the-art organic nonlinear exhibiting in-plane axis.

Multi-agent deep reinforcement learning has made a great achievement in through modeling real-world scenario with multiple agents that communicate single environment. However, the test and validation of MARL model on conventional multi-agent simulation are limited. In this study, we analyze an effective method to use validate models methods as well propose two requirements, intuitive interface optimization simulation, achieve it.

In this study, we developed a three-dimensional (3D) in vitro infection model to investigate the crosstalk between phagocytes and microbes inflammation using nanofibrous membrane (NM). Poly(ε-caprolactone) (PCL)-NMs (PCL-NMs) were generated via electrospinning of PCL chloroform. Staphylococcus aureus able adhere nanofibers engulfed S. PCL-NM. The migration was evaluated two-layer co-culture system Neutrophils, macrophages dendritic cells (DCs) cultured upper PCL-NM layer migrated lower...

Vision Transformers (ViTs) have achieved remarkable success in various computer vision tasks. However, ViTs a huge computational cost due to their inherent reliance on multi-head self-attention (MHSA), prompting efforts accelerate for practical applications. To this end, recent works aim reduce the number of tokens, mainly focusing how effectively prune or merge them. Nevertheless, since ViT tokens are generated from non-overlapping grid patches, they usually do not convey sufficient...

Federated learning (FL) was proposed for training a deep neural network model using millions of user data. The technique has attracted considerable attention owing to its privacy-preserving characteristic. However, two major challenges exist. first is the limitation simultaneously participating clients. If number clients increases, single parameter server easily becomes bottleneck and prone have stragglers. second data heterogeneity, which adversely affects accuracy global model. Because...

Given a convex polygon $P$ with $k$ vertices and polygonal domain $Q$ consisting of obstacles total size $n$ in the plane, we study optimization problem finding largest similar copy that can be placed without intersecting obstacles. We improve time complexity for solving to $O(k^2n^2λ_4(k)\log{n})$. This is progress improving previously best known results by Chew Kedem [SoCG89, CGTA93] Sharir Toledo [SoCG91, CGTA94] on more than 25 years.

Solving partial differential equations (PDEs) by learning the solution operators has emerged as an attractive alternative to traditional numerical methods. However, implementing such architectures presents two main challenges: flexibility in handling irregular and arbitrary input output formats scalability large discretizations. Most existing are limited their desired structure or infeasible scale inputs outputs. To address these issues, we introduce attention-based model called...

In article number 1605583, Fabian Rotermund, O-Pil Kwon, and co-workers present new organic electro-optic quinolinium salt crystals exhibiting high optical-to-terahertz (THz) conversion efficiencies. Large optical nonlinearity lower THz absorption form the basis for this efficiency. This is achieved by suppressing vibrational modes introduction of strongly electronegative groups acting as strong hydrogen-bond acceptor sites into molecular anions.

We study the problem of finding maximum-area rectangles contained in a polygon plane. There has been fair amount work for this when have to be axis-aligned or is convex. consider simple with $n$ vertices, possibly holes, and no restriction on orientation rectangles. present an algorithm that computes rectangle $O(n^3\log n)$ time using $O(kn^2)$ space, where $k$ number reflex vertices $P$. Our can report all same $O(n^3)$ space. also finds convex $O(n)$

When a physical object is scanned with an optical 3D scanner in normal environment, shading caused by external lighting that effect on the color data. This not easy to separate from original surface of model and can make abnormal rendering result when rendered additional artificial lighting. Therefore, various kind automatic or manual processing required suppressing general. However, requires information environment which simple sense record. In our research, using white spherical plaster as...

We study the problem of finding maximum-area triangles that can be inscribed in a polygon plane. consider eight versions problem: we use either convex polygons or simple as container; require to have one corner with fixed angle all three corners angles; allow reorienting triangle its orientation fixed. present exact algorithms for problem. In case reorientations $n$ vertices, also $(1-\varepsilon)$-approximation algorithms.

Given a convex polygon with k vertices and polygonal domain consisting of obstacles n in total the plane, we study optimization problem finding largest similar copy that can be placed without intersecting obstacles. We present an upper bound O ( 2 λ 4 )) on number combinatorial changes occurred to underlying structure during rotation polygon, together )log )-time deterministic algorithm for problem. This improves upon previously best known results by Chew Kedem [SoCG89, CGTA93] Sharir Toledo...

First fit decreasing (FFD) is the most popular heuristic for virtual machine (VM) placement problems. However, FFD does not perform as much in a heterogeneous cluster environment. Moreover, and other heuristics, such best (BFD), are limited to handle VM problem effectively when multiple resources considered together. In this study, we analyze reason why ant colony system performs better than cluster. We verified our logical observations through experimental comparisons with heuristics.