Showing papers by "Nanyang Technological University published in 2022"

Aspect-based sentiment analysis via affective knowledge enhanced graph convolutional networks

Abstract: Aspect-based sentiment analysis is a fine-grained sentiment analysis task, which needs to detection the sentiment polarity towards a given aspect. Recently, graph neural models over the dependency tree are widely applied for aspect-based sentiment analysis. Most existing works, however, they generally focus on learning the dependency information from contextual words to aspect words based on the dependency tree of the sentence, which lacks the exploitation of contextual affective knowledge with regard to the specific aspect. In this paper, we propose a graph convolutional network based on SenticNet to leverage the affective dependencies of the sentence according to the specific aspect, called Sentic GCN. To be specific, we explore a novel solution to construct the graph neural networks via integrating the affective knowledge from SenticNet to enhance the dependency graphs of sentences. Based on it, both the dependencies of contextual words and aspect words and the affective information between opinion words and the aspect are considered by the novel affective enhanced graph model. Experimental results on multiple public benchmark datasets illustrate that our proposed model can beat state-of-the-art methods.

Decentralized Edge Intelligence: A Dynamic Resource Allocation Framework for Hierarchical Federated Learning

Abstract: To enable the large scale and efficient deployment of Artificial Intelligence (AI), the confluence of AI and Edge Computing has given rise to Edge Intelligence, which leverages on the computation and communication capabilities of end devices and edge servers to process data closer to where it is produced. One of the enabling technologies of Edge Intelligence is the privacy preserving machine learning paradigm known as Federated Learning (FL), which enables data owners to conduct model training without having to transmit their raw data to third-party servers. However, the FL network is envisioned to involve thousands of heterogeneous distributed devices. As a result, communication inefficiency remains a key bottleneck. To reduce node failures and device dropouts, the Hierarchical Federated Learning (HFL) framework has been proposed whereby cluster heads are designated to support the data owners through intermediate model aggregation. This decentralized learning approach reduces the reliance on a central controller, e.g., the model owner. However, the issues of resource allocation and incentive design are not well-studied in the HFL framework. In this article, we consider a two-level resource allocation and incentive mechanism design problem. In the lower level, the cluster heads offer rewards in exchange for the data owners’ participation, and the data owners are free to choose which cluster to join. Specifically, we apply the evolutionary game theory to model the dynamics of the cluster selection process. In the upper level, each cluster head can choose to serve a model owner, whereas the model owners have to compete amongst each other for the services of the cluster heads. As such, we propose a deep learning based auction mechanism to derive the valuation of each cluster head's services. The performance evaluation shows the uniqueness and stability of our proposed evolutionary game, as well as the revenue maximizing properties of the deep learning based auction.

Guiding Transition Metal‐Doped Hollow Cerium Tandem Nanozymes with Elaborately Regulated Multi‐Enzymatic Activities for Intensive Chemodynamic Therapy

Abstract: Clinical applications of nanozyme-initiated chemodynamic therapy (NCDT) have been severely limited by the poor catalytic efficiency of nanozymes, insufficient endogenous hydrogen peroxide (H2 O2 ) content, and its off-target consumption. Herein, the authors developed a hollow mesoporous Mn/Zr-co-doped CeO2 tandem nanozyme (PHMZCO-AT) with regulated multi-enzymatic activities, that is, the enhancement of superoxide dismutase (SOD)-like and peroxidase (POD)-like activities and inhibition of catalase (CAT)-like activity. PHMZCO-AT as a H2 O2 homeostasis disruptor promotes H2 O2 evolution and restrains off-target elimination of H2 O2 to achieve intensive NCDT. PHMZCO-AT with SOD-like activity catalyzes endogenous superoxide anion (O2•- ) into H2 O2 in the tumor region. The suppression of CAT activity and depletion of glutathione by PHMZCO-AT largely weaken the off-target decomposition of H2 O2 to H2 O. Elevated H2 O2 is then catalyzed by the downstream POD-like activity of PHMZCO-AT to generate toxic hydroxyl radicals, further inducing tumor apoptosis and death. T1 -weighted magnetic resonance imaging and X-ray computed tomography imaging are also achieved using PHMZCO-AT due to the existence of paramagnetic Mn2+ and the high X-ray attenuation ability of elemental Zr, permitting in vivo tracking of the therapeutic process. This work presents a typical paradigm to achieve intensive NCDT efficacy by regulating multi-enzymatic activities of nanozymes to perturb the H2 O2 homeostasis.

The Chemistry of Organic Contrast Agents in the NIR‐II Window

Abstract: Optical imaging, especially fluorescence and photoacoustic imaging, is a non-invasive imaging approach with high spatial and temporal resolution and high sensitivity, compared to positron emission tomography (PET) or magnetic resonance imaging (MRI). Due to the merits of imaging using the second near-infrared (NIR-II) window, like deeper penetration depth, high signal-to-noise ratio, high resolution, and low tissue damage, researchers have devoted great effort to developing contrast agents with NIR-II absorption or emission. In this Review, we summarize recently developed organic luminescent and photoacoustic materials, ranging from small molecules to conjugated polymers. Then, we systematically introduce engineering strategies and describe the imaging performance classified by the skeleton cores. Finally, we elucidate the challenges and prospects of these NIR-II organic dyes for potential clinical applications. We expect our summary can inspire researchers to expand the spectrum of NIR-II contrast agents for diverse bioapplications.

Activatable molecular probes for fluorescence-guided surgery, endoscopy and tissue biopsy

Abstract: The real-time, dynamic optical visualization of lesions and margins ensures not only complete resection of the malignant tissues but also better preservation of the vital organs/tissues during surgical procedures. Most imaging probes with an "always-on" signal encounter high background noise due to their non-specific accumulation in normal tissues. By contrast, activatable molecular probes only "turn on" their signals upon reaction with the targeted biomolecules that are overexpressed in malignant cells, offering high target-to-background ratios with high specificity and sensitivity. This review summarizes the recent progress of activatable molecular probes in surgical imaging and diagnosis. The design principle and mechanism of activatable molecular probes are discussed, followed by specific emphasis on applications ranging from fluorescence-guided surgery to endoscopy and tissue biopsy. Finally, potential challenges and perspectives in the field of activatable molecular probe-enabled surgical imaging are discussed.

Chemical recycling of plastic waste for sustainable material management: A prospective review on catalysts and processes

Abstract: Plastic waste management becomes imperative to prevent detrimental plastic-related environmental impacts. These relate to the huge plastic waste amount in both generation and disposal, non-biodegradability of most plastic waste, persistent nature of plastic waste in the environment, and the increasing concern of micro- and nanoplastics on human health. To continue exporting the use and benefits of plastic materials while mitigating their impacts to the planet, reuse, recovery and recycling of plastic waste are strongly encouraged. This would promote optimization of the value chain of plastic-based materials and validate the concept of a circular polymer economy. Chemical recycling techniques have been extensively studied for energy and material recovery from plastic waste, which are assisted by catalysts for enhanced process efficiency. In this review, the research progress of the applied catalysts in various thermochemical techniques is comprehensively summarized and assessed, covering the non-catalytic and catalytic pyrolysis, reforming of pyrolysis volatiles, and gasification processes. The impacts of feedstock, process conditions and catalyst properties on the quality and yield of the products are carefully examined. Finally, the practical challenges and perspectives on catalyst design and process improvement for material circularity are suggested and the emerging novel processes for the degradation of plastic waste are highlighted.

Chemical recycling of plastic waste for sustainable material management: A prospective review on catalysts and processes

Abstract: Plastic waste management becomes imperative to prevent detrimental plastic-related environmental impacts. These relate to the huge plastic waste amount in both generation and disposal, non-biodegradability of most plastic waste, persistent nature of plastic waste in the environment, and the increasing concern of micro- and nanoplastics on human health. To continue exporting the use and benefits of plastic materials while mitigating their impacts to the planet, reuse, recovery and recycling of plastic waste are strongly encouraged. This would promote optimization of the value chain of plastic-based materials and validate the concept of a circular polymer economy. Chemical recycling techniques have been extensively studied for energy and material recovery from plastic waste, which are assisted by catalysts for enhanced process efficiency. In this review, the research progress of the applied catalysts in various thermochemical techniques is comprehensively summarized and assessed, covering the non-catalytic and catalytic pyrolysis, reforming of pyrolysis volatiles, and gasification processes. The impacts of feedstock, process conditions and catalyst properties on the quality and yield of the products are carefully examined. Finally, the practical challenges and perspectives on catalyst design and process improvement for material circularity are suggested and the emerging novel processes for the degradation of plastic waste are highlighted.

Strategies for enhancing cancer chemodynamic therapy performance

Abstract: Chemodynamic therapy (CDT) has emerged to be a frontrunner amongst reactive oxygen species-based cancer treatment modalities. CDT utilizes endogenous H2O2 in tumor microenvironment (TME) to produce cytotoxic hydroxyl radicals (•OH) via Fenton or Fenton-like reactions. While possessing advantages such as tumor specificity, no need of external stimuli, and low side effects, practical applications of CDT are still impeded owing to the heterogeneity, complexity, and reductive environment of TME. Over the past couple of years, strategies to enhance CDT for efficient tumor regression are in rapid development in synergy with the growth of nanomedicine. In this review, we initially outline the fundamental understanding of Fenton and Fenton-like reactions and their relationship with CDT. Subsequently, the development in the design of nanosystems for CDT is highlighted in a general manner. Furthermore, recent advancement of the strategies to augment Fenton reactions in TME for enhanced CDT is discussed in detail. Finally, perspectives toward the future development of CDT for better therapeutic outcome are presented. This review is expected to draw attention for collaborative research on CDT in the best interest of its future clinical applications.

State-of-Health Estimation of Lithium-Ion Batteries by Fusing an Open Circuit Voltage Model and Incremental Capacity Analysis

Abstract: The state of health (SOH) is a vital parameter enabling the reliability and life diagnostic of lithium-ion batteries. A novel fusion-based SOH estimator is proposed in this study, which combines an open circuit voltage (OCV) model and the incremental capacity analysis. Specifically, a novel OCV model is developed to extract the OCV curve and the associated features-of-interest (FOIs) from the measured terminal voltage during constant-current charge. With the determined OCV model, the disturbance-free incremental capacity (IC) curves can be derived, which enables the extraction of a set of IC morphological FOIs. The extracted model FOI and IC morphological FOIs are further fused for SOH estimation through an artificial neural network. Long-term degradation data obtained from different battery chemistries are used for validation. Results suggest that the proposed fusion-based method manifests itself with high estimation accuracy and high robustness.

Data-Driven Estimation of Driver Attention Using Calibration-Free Eye Gaze and Scene Features

Abstract: Driver attention estimation is one of the key technologies for intelligent vehicles The existing related methods only focus on the scene image or the driver's gaze or head pose The purpose of this article is to propose a more reasonable and feasible method based on a dual-view scene with calibration-free gaze direction According to human visual mechanisms, the low-level features, static visual saliency map, and dynamic optical flow information are extracted as input feature maps, which combine the high-level semantic descriptions and a gaze probability map transformed from the gaze direction A multiresolution neural network is proposed to handle the calibration-free features The proposed method is verified on a virtual reality experimental platform that collected more than 550 000 samples and obtained a more accurate ground truth The experiments show that the proposed method is feasible and better than the state-of-the-art methods based on multiple widely used metrics This study also provides a discussion of the effects of different landscapes, times, and weather conditions on the performance

Design Methodology of Free-Positioning Nonoverlapping Wireless Charging for Consumer Electronics Based on Antiparallel Windings

Abstract: Free positioning wireless charging for consumer electronics allows the devices to be charged at arbitrary positions and angles to improve user experience. However, a user-initiated sudden movement of the device during charging can cause hazards due to the abrupt variation of the coupling coefficient. To solve this issue, the coupling coefficient variation at different positions should be mitigated, which is also good for the design and high-efficiency operation of power electronics converters. This article proposes a design methodology to employ antiparallel windings to smooth the coupling coefficient variation over different positions by reducing the coupling coefficient at central positions and enhancing it at boundary positions. Two optimization methods are proposed: turn-by-turn optimization and winding-by-winding optimization. A design flow is offered. The hexagonal coil is compared with the square coil and proved to achieve better performance than the latter. An experimental prototype is implemented to validate the effectiveness of the proposed design methodology.

Interface formation and deformation behaviors of an additively manufactured nickel-aluminum-bronze/15-5 PH multimaterial via laser-powder directed energy deposition

Abstract: Additive manufacturing (AM) of a nickel-aluminum-bronze (NAB)/15-5 PH multimaterial by laser-powder directed energy deposition (LP-DED) accomplished a combination of excellent mechanical performance and high corrosion resistance. An NAB/15-5 PH interface without cracks and lack of fusion was achieved, which was characterized with an interlayer of FexAl dendrites. The formation of the interfacial characteristics was attributed to a synthetic effect of liquid phase separation, Marangoni convection, and atom diffusion. A miscibility gap was generated by a high degree of supercooling in the melt pool, and 15-5 PH solidified prior to NAB to form a dendritic interlayer. Marangoni convection occurred to promote the Al atom diffusion from NAB to 15-5 PH, contributing to the formation of the FexAl phase at the interface. The multimaterial sample possessed higher ultimate tensile strength of 754.64 MPa in the transverse direction and 854.57 MPa in the longitudinal direction as compared to that of copper/steel counterparts fabricated by AM. The multimaterial printed by LP-DED exhibited different deformation mechanisms in the transverse and longitudinal directions. In the transverse direction, NAB contributed more deformation than 15-5 PH and determined the improved ductility of the multimaterial; in the longitudinal direction, the brittle FexAl dendrites constrained the deformation of NAB and 15-5 PH, which resulted in the early failure of the multimaterial. The multimaterial tended to undergo cracking at the interface of the FexAl and Cu phases under stress concentration, which was induced by their crystal incoherence.

Tumor‐Microenvironment‐Activatable Polymer Nano‐Immunomodulator for Precision Cancer Photoimmunotherapy

Abstract: Cancer nanomedicine combined with immunotherapy has become a promising strategy for treating cancer in terms of safety and potency; however, precise regulation of the activation of antitumor immunity remains challenging. Herein, a smart semiconducting polymer nano-immunomodulator (SPNI), which responds to the acidic tumor microenvironment (TME), for precision photodynamic immunotherapy of cancer, is reported. The SPNI is self-assembled by a near-infrared (NIR)-absorbing semiconducting polymer and an amphipathic polymer conjugated with a Toll-like receptor 7 (TLR7) agonist via an acid-labile linker. Upon arrival at tumor site, SPNI undergoes hydrolysis and triggers an efficient liberation of TLR7 agonist in response to the acidic TME for dendritic cell activation. Moreover, SPNI exerts photodynamic effects for direct tumor eradication and immunogenic cancer cell death under NIR photoirradiation. The synergistic action of released immunogenic factors and acidic-TME-activated TLR7 agonist can serve as an in situ generated cancer vaccine to evoke strong antitumor activities. Notably, such localized immune activation boosts systemic antitumor immune responses, resulting in enhanced cytotoxic CD8+ T infiltration to inhibit tumor growth and metastasis. Thereby, this work presents a general strategy to devise prodrug of immunotherapeutics for precise regulation of cancer immunotherapy.