A p-vinylbenzyl sulfobetaine was grafted from cellulose membrane (CM) using surface-initiated atom transfer radical polymerization for blood compatibility improvement. Surface structure, wettability, morphology, and thermal stability of the CM substrates before after modification were characterized by attenuated total reflectance Fourier transform infrared spectra, X-ray photoelectron spectroscopy measurement, water contact angle atomic force microscopy, thermogravimetric analysis,...

Nearshore multi-object tracking (NMOT) aims to locat and identify nearshore objects. Most approaches accomplish this task using radar remote-sensing technologies. In contrast, video data can describe the visual appearance of objects without prior information, such as identity, location, or movement. study, we introduce a cross-modality pipeline address four major challenges NMOT. First, propose introducing bi-attention transformer (CBT) manage information interaction between RGB thermal...

In this work, two kinds of starch-based flocculants with the same chemically modified functional groups but far different substitution degrees, (2-hydroxypropyl)trimethylammonium chloride etherified carboxymethyl starch (denoted as CMS-CTA-P and CMS-CTA-N, respectively) were successfully prepared. CMS-CTA-N bear opposite surface charge properties in water under most measured pH range. The flocculation performance these has been systematically studied using kaolin hematite suspensions...

The world has seen in recent years great successes applying deep learning (DL) for many application domains. Though powerful, DL is not easy to be used well. In this invited paper, we study an urban taxi demand forecast problem using DL, and show a number of key insights modeling domain as suitable task. We also conduct systematic comparison two neural networks (DNNs) prediction, i.s., the ST-ResNet FLC-Net, on New York city record dataset. Our experimental results DNNs indeed outperform...

The Internet of Things (IoT) and blockchain, the hottest frontier technologies in recent years, are expected to lead next technological revolution. Blockchain promises solve current challenges encountered by IoT. However, most proposed blockchain-based IoT architectures, which based on discrete logarithm or large integer factorization problems, susceptible quantum attacks. Several quantum-resistant blockchain schemes have recently been proposed. efficiency their construction equipment...

Existing multi-class text classification techniques prevailingly adhere to a uniform training paradigm that consists of an encoder for encoding, linear classifier label prediction, and one-hot vectors representation. However, this conventional considers labels as independent categorical variables, overlooking the implicit semantic information contained within labels. This may result in models exhibiting overconfidence, particularly scenarios with semantically similar or inaccurately labeled...

AB diblock copolymers comprised of poly(2-methacryloyloxyethyl phosphorylcholine) (PMPC) and poly(3-methacryloxypropyl trimethoxysilane) (PMTSi) segments, which are used for biocompatible coatings, were investigated. Block with various compositions synthesized by atomic transfer radical polymerization (ATRP). The obtained dissolved in an ethanol solution, dynamic light scattering showed that all block capable existing as micelles. After a convenient "one-step" reaction, the cellulose...

Poly(3-trimethoxysilyl-propyl-methacrylate) (PMPS) and poly(2-methacryloyloxyethyl phosphorylcholine) (PMPC) with various compositions was synthesized for the surface modification of currently commercially available lenses. The properties these anti-biofouling contact lenses were investigated, which indicated that modified had significant surface-wettability improvement, superior antibacterial adhesion resistance to protein adsorption. PMPS-b-PMPC can be immobilized on by one-step while...

Continual learning, also known as incremental learning or life-long stands at the forefront of deep and AI systems. It breaks through obstacle one-way training on close sets enables continuous adaptive open-set conditions. In recent decade, continual has been explored applied in multiple fields especially computer vision covering classification, detection segmentation tasks. semantic (CSS), which dense prediction peculiarity makes it a challenging, intricate burgeoning task. this paper, we...

Visual-language generation for remote sensing image (RSI) is an emerging and challenging research area that requires multi-task learning to achieve a comprehensive understanding. However, most existing models are limited single-level tasks do not leverage the advantages of Visual-Language Pre-training (VLP) model. In this paper, we present unified benchmark learns multiple tasks, including interpretation, perception, question answering. Specifically, model designed perform semantic...

Current remote-sensing interpretation models often focus on a single task such as detection, segmentation, or caption. However, the task-specific designed are unattainable to achieve comprehensive multi-level of images. The field also lacks support for multi-task joint datasets. In this paper, we propose Panoptic Perception: novel and new fine-grained dataset (FineGrip) more thorough universal RSIs. task: 1) integrates pixel-level, instance-level, image-level information image perception, 2)...