Search

Local geometric information, i.e. normal and point distribution, is crucial for LiDAR-based simultaneous localization and mapping (SLAM) because it provides constrains for data association, which further determines the direction of optimization and ultimately affects the accuracy of poses. However, estimating normal and point distribution are time-consuming tasks even with the assistance of the KDtree or volumetic maps. To achieve fast normal estimation, we look into the structural information of LiDAR scan and propose a novel fast approximate least squares (FALS) method. With the pre-computed bearing information, estimating the normal requires only the range information of the points when a new scan arrives. To efficiently estimate the distribution of points, we extend the ikd-tree to manage the map in voxels and update its point cloud distribution incrementally while maintaining its consistency with the normals. For scan points that satisfy visibility and consistency checks based on normal, we devise a robust and accurate hierarchical data association schema considering the distribution where point-to-surfel is prioritized over point-to-plane. We further fix voxels after the distribution convergences to balance the time consumption and the correctness of representation. Extensive experiments on diverse public datasets demonstrate the advantages of our system compared to other state-of-the-art methods., 8 pages, 4 figures

Continuously-worn wearable sensors enable researchers to collect copious amounts of rich bio-behavioral time series recordings of real-life activities of daily living, offering unprecedented opportunities to infer novel human behavior patterns during daily routines. Existing approaches to routine discovery through bio-behavioral data rely either on pre-defined notions of activities or use additional non-behavioral measurements as contexts, such as GPS location or localization within the home, presenting risks to user privacy. In this work, we propose a novel wearable time-series mining framework, Hawkes point process On Time series clusters for ROutine Discovery (HOT-ROD), for uncovering behavioral routines from completely unlabeled wearable recordings. We utilize a covariance-based method to generate time-series clusters and discover routines via the Hawkes point process learning algorithm. We empirically validate our approach for extracting routine behaviors using a completely unlabeled time-series collected continuously from over 100 individuals both in and outside of the workplace during a period of ten weeks. Furthermore, we demonstrate this approach intuitively captures daily transitional relationships between physical activity states without using prior knowledge. We also show that the learned behavioral patterns can assist in illuminating an individual's personality and affect., 2023 9th ACM SIGKDD International Workshop on Mining and Learning From Time Series (MiLeTS 2023)

Over the past few years, Federated Learning (FL) has become an emerging machine learning technique to tackle data privacy challenges through collaborative training. In the Federated Learning algorithm, the clients submit a locally trained model, and the server aggregates these parameters until convergence. Despite significant efforts that have been made to FL in fields like computer vision, audio, and natural language processing, the FL applications utilizing multimodal data streams remain largely unexplored. It is known that multimodal learning has broad real-world applications in emotion recognition, healthcare, multimedia, and social media, while user privacy persists as a critical concern. Specifically, there are no existing FL benchmarks targeting multimodal applications or related tasks. In order to facilitate the research in multimodal FL, we introduce FedMultimodal, the first FL benchmark for multimodal learning covering five representative multimodal applications from ten commonly used datasets with a total of eight unique modalities. FedMultimodal offers a systematic FL pipeline, enabling end-to-end modeling framework ranging from data partition and feature extraction to FL benchmark algorithms and model evaluation. Unlike existing FL benchmarks, FedMultimodal provides a standardized approach to assess the robustness of FL against three common data corruptions in real-life multimodal applications: missing modalities, missing labels, and erroneous labels. We hope that FedMultimodal can accelerate numerous future research directions, including designing multimodal FL algorithms toward extreme data heterogeneity, robustness multimodal FL, and efficient multimodal FL. The datasets and benchmark results can be accessed at: https://github.com/usc-sail/fed-multimodal., This paper was accepted to KDD 2023 Applied Data Science (ADS) track

Many recent studies have focused on fine-tuning pre-trained models for speech emotion recognition (SER), resulting in promising performance compared to traditional methods that rely largely on low-level, knowledge-inspired acoustic features. These pre-trained speech models learn general-purpose speech representations using self-supervised or weakly-supervised learning objectives from large-scale datasets. Despite the significant advances made in SER through the use of pre-trained architecture, fine-tuning these large pre-trained models for different datasets requires saving copies of entire weight parameters, rendering them impractical to deploy in real-world settings. As an alternative, this work explores parameter-efficient fine-tuning (PEFT) approaches for adapting pre-trained speech models for emotion recognition. Specifically, we evaluate the efficacy of adapter tuning, embedding prompt tuning, and LoRa (Low-rank approximation) on four popular SER testbeds. Our results reveal that LoRa achieves the best fine-tuning performance in emotion recognition while enhancing fairness and requiring only a minimal extra amount of weight parameters. Furthermore, our findings offer novel insights into future research directions in SER, distinct from existing approaches focusing on directly fine-tuning the model architecture. Our code is publicly available under: https://github.com/usc-sail/peft-ser., This work was accepted to the 11th International Conference on Affective Computing and Intelligent Interaction (ACII), 2023

In this work, we propose GPT-FL, a generative pre-trained model-assisted federated learning (FL) framework. At its core, GPT-FL leverages generative pre-trained models to generate diversified synthetic data. These generated data are used to train a downstream model on the server, which is then fine-tuned with private client data under the standard FL framework. We show that GPT-FL consistently outperforms state-of-the-art FL methods in terms of model test accuracy, communication efficiency, and client sampling efficiency. Through comprehensive ablation analysis, we discover that the downstream model generated by synthetic data plays a crucial role in controlling the direction of gradient diversity during FL training, which enhances convergence speed and contributes to the notable accuracy boost observed with GPT-FL. Also, regardless of whether the target data falls within or outside the domain of the pre-trained generative model, GPT-FL consistently achieves significant performance gains, surpassing the results obtained by models trained solely with FL or synthetic data.

Speech processing techniques are useful for analyzing speech and language development in children with Autism Spectrum Disorder (ASD), who are often varied and delayed in acquiring these skills. Early identification and intervention are crucial, but traditional assessment methodologies such as caregiver reports are not adequate for the requisite behavioral phenotyping. Natural Language Sample (NLS) analysis has gained attention as a promising complement. Researchers have developed benchmarks for spoken language capabilities in children with ASD, obtainable through the analysis of NLS. This paper proposes applications of speech processing technologies in support of automated assessment of children's spoken language development by classification between child and adult speech and between speech and nonverbal vocalization in NLS, with respective F1 macro scores of 82.6% and 67.8%, underscoring the potential for accurate and scalable tools for ASD research and clinical use., Accepted to Interspeech 2023, 5 pages

Recent studies have explored the use of pre-trained embeddings for speech emotion recognition (SER), achieving comparable performance to conventional methods that rely on low-level knowledge-inspired acoustic features. These embeddings are often generated from models trained on large-scale speech datasets using self-supervised or weakly-supervised learning objectives. Despite the significant advancements made in SER through the use of pre-trained embeddings, there is a limited understanding of the trustworthiness of these methods, including privacy breaches, unfair performance, vulnerability to adversarial attacks, and computational cost, all of which may hinder the real-world deployment of these systems. In response, we introduce TrustSER, a general framework designed to evaluate the trustworthiness of SER systems using deep learning methods, with a focus on privacy, safety, fairness, and sustainability, offering unique insights into future research in the field of SER. Our code is publicly available under: https://github.com/usc-sail/trust-ser.

Automatic Speech Understanding (ASU) leverages the power of deep learning models for accurate interpretation of human speech, leading to a wide range of speech applications that enrich the human experience. However, training a robust ASU model requires the curation of a large number of speech samples, creating risks for privacy breaches. In this work, we investigate using foundation models to assist privacy-enhancing speech computing. Unlike conventional works focusing primarily on data perturbation or distributed algorithms, our work studies the possibilities of using pre-trained generative models to synthesize speech content as training data with just label guidance. We show that zero-shot learning with training label-guided synthetic speech content remains a challenging task. On the other hand, our results demonstrate that the model trained with synthetic speech samples provides an effective initialization point for low-resource ASU training. This result reveals the potential to enhance privacy by reducing user data collection but using label-guided synthetic speech content.

As the most sensitive key area of the Earth system to global climate change, research on the key processes of the cryosphere has attracted much attention. Meanwhile, the inevitable trend in the multi-scale monitoring study of key processes of the Earth system is to move from digital surveying to intelligent surveying and mapping. This article summarizes the research status quo on the key processes of the Earth system in terms of intelligent observation and processing in five aspects:glacier melting and sea level rise, polar ice cap stability and subsurface structure, Arctic and Antarctic sea ice changes and extreme climates, permafrost degradation and geological disasters, and livable urban underground spaces under the Earth system. Furthermore, it provides the outlook of the development trend of the intelligent surveying and mapping of key areas and key processes of the Earth system, i.e., to improve the intelligent comprehensive monitoring network, to establish big data centers for key areas, and to build the intelligent model and prediction system of key processes.

Visual Place Recognition (VPR) in areas with similar scenes such as urban or indoor scenarios is a major challenge. Existing VPR methods using global descriptors have difficulty capturing local specific regions (LSR) in the scene and are therefore prone to localization confusion in such scenarios. As a result, finding the LSR that are critical for location recognition becomes key. To address this challenge, we introduced Patch-NetVLAD+, which was inspired by patch-based VPR researches. Our method proposed a fine-tuning strategy with triplet loss to make NetVLAD suitable for extracting patch-level descriptors. Moreover, unlike existing methods that treat all patches in an image equally, our method extracts patches of LSR, which present less frequently throughout the dataset, and makes them play an important role in VPR by assigning proper weights to them. Experiments on Pittsburgh30k and Tokyo247 datasets show that our approach achieved up to 6.35\% performance improvement than existing patch-based methods.

The national carbon emission trading mechanism is an important policy tool for the Chinese government to control and reduce greenhouse gas emissions by using the market mechanism. The Beijing–Tianjin–Hebei power market is the focus of energy conservation and consumption reduction in China. Problems have already existed in the synergic development of the Beijing–Tianjin–Hebei power market and carbon trading market. In this article, the development status of the Beijing–Tianjin–Hebei power market is analyzed and the coupling mechanism between the carbon market and power market is combed out to build a synergism model of the carbon market and the Beijing–Tianjin–Hebei power market based on the system dynamics. From the research results, firstly, the Beijing–Tianjin–Hebei power market comes with a high energy consumption intensity and a high proportion of carbon emissions. The coupling of carbon market and power market forces the power industry to reduce carbon emissions through the effective transmission of carbon costs to power prices. Secondly, carbon price shows an upward trend in the context of the current policy scenario, which can give play to the role of price signal in the future. The revenue of thermal power plants, which are the carbon emission right sellers, with new technologies, has increased significantly, while the revenue of carbon emission right buyers, which are the manufacturers of undeveloped units, has increased less. Finally, the technical progress of thermal power plants, the introduction of auction mechanism, the increase in initial carbon price settings and the direct transmission of carbon costs are all factors that promote the effectiveness of carbon trading policy tools in the Beijing–Tianjin–Hebei power market. This study provides theoretical guidance for the synergic development of the “power-carbon” market.

Interpersonal spoken communication is central to human interaction and the exchange of information. Such interactive processes involve not only speech and spoken language but also non-verbal cues such as hand gestures, facial expressions, and nonverbal vocalization, that are used to express feelings and provide feedback. These multimodal communication signals carry a variety of information about the people: traits like gender and age as well as about physical and psychological states and behavior. This work uses wearable multimodal sensors to investigate interpersonal communication behaviors focusing on speaking patterns among healthcare providers with a focus on nurses. We analyze longitudinal data collected from $99$ nurses in a large hospital setting over ten weeks. The results indicate that speaking pattern differences across shift schedules and working units. Moreover, results show that speaking patterns combined with physiological measures can be used to predict affect measures and life satisfaction scores. The implementation of this work can be accessed at https://github.com/usc-sail/tiles-audio-arousal.

Speech-centric machine learning systems have revolutionized many leading domains ranging from transportation and healthcare to education and defense, profoundly changing how people live, work, and interact with each other. However, recent studies have demonstrated that many speech-centric ML systems may need to be considered more trustworthy for broader deployment. Specifically, concerns over privacy breaches, discriminating performance, and vulnerability to adversarial attacks have all been discovered in ML research fields. In order to address the above challenges and risks, a significant number of efforts have been made to ensure these ML systems are trustworthy, especially private, safe, and fair. In this paper, we conduct the first comprehensive survey on speech-centric trustworthy ML topics related to privacy, safety, and fairness. In addition to serving as a summary report for the research community, we point out several promising future research directions to inspire the researchers who wish to explore further in this area.

Speech emotion recognition (SER) processes speech signals to detect and characterize expressed perceived emotions. Many SER application systems often acquire and transmit speech data collected at the client-side to remote cloud platforms for inference and decision making. However, speech data carry rich information not only about emotions conveyed in vocal expressions, but also other sensitive demographic traits such as gender, age and language background. Consequently, it is desirable for SER systems to have the ability to classify emotion constructs while preventing unintended/improper inferences of sensitive and demographic information. Federated learning (FL) is a distributed machine learning paradigm that coordinates clients to train a model collaboratively without sharing their local data. This training approach appears secure and can improve privacy for SER. However, recent works have demonstrated that FL approaches are still vulnerable to various privacy attacks like reconstruction attacks and membership inference attacks. Although most of these have focused on computer vision applications, such information leakages exist in the SER systems trained using the FL technique. To assess the information leakage of SER systems trained using FL, we propose an attribute inference attack framework that infers sensitive attribute information of the clients from shared gradients or model parameters, corresponding to the FedSGD and the FedAvg training algorithms, respectively. As a use case, we empirically evaluate our approach for predicting the client's gender information using three SER benchmark datasets: IEMOCAP, CREMA-D, and MSP-Improv. We show that the attribute inference attack is achievable for SER systems trained using FL. We further identify that most information leakage possibly comes from the first layer in the SER model.

Estimating geometric elements such as depth, camera motion, and optical flow from images is an important part of the robot's visual perception. We use a joint self-supervised method to estimate the three geometric elements. Depth network, optical flow network and camera motion network are independent of each other but are jointly optimized during training phase. Compared with independent training, joint training can make full use of the geometric relationship between geometric elements and provide dynamic and static information of the scene. In this paper, we improve the joint self-supervision method from three aspects: network structure, dynamic object segmentation, and geometric constraints. In terms of network structure, we apply the attention mechanism to the camera motion network, which helps to take advantage of the similarity of camera movement between frames. And according to attention mechanism in Transformer, we propose a plug-and-play convolutional attention module. In terms of dynamic object, according to the different influences of dynamic objects in the optical flow self-supervised framework and the depth-pose self-supervised framework, we propose a threshold algorithm to detect dynamic regions, and mask that in the loss function respectively. In terms of geometric constraints, we use traditional methods to estimate the fundamental matrix from the corresponding points to constrain the camera motion network. We demonstrate the effectiveness of our method on the KITTI dataset. Compared with other joint self-supervised methods, our method achieves state-of-the-art performance in the estimation of pose and optical flow, and the depth estimation has also achieved competitive results. Code will be available https://github.com/jianfenglihg/Unsupervised_geometry., 9 pages, 4 figures

Two-dimensional polyimide covalent organic framework (2D PI-NT COF) films were constructed on indium tin oxide-coated glass substrates to fabricate two-terminal sandwiched resistive memory devices. The 2D PI-NT COF films condensated from the reaction between 4,4',4″-triaminotriphenylamine and naphthalene-1,4,5,8-tetracarboxylic dianhydride under solvothermal conditions demonstrated high crystallinity, good orientation preference, tunable thickness, and low surface roughness. The well-aligned electron-donor (triphenylamine unit) and -acceptor (naphthalene diimide unit) arrays rendered the 2D PI-NT COF films a promising candidate for electronic applications. The memory devices based on 2D PI-NT COF films exhibited a typical write-once-read-many-time resistive switching behavior under an operating voltage of +2.30 V on the positive scan and -2.64 V on the negative scan. A high ON/OFF current ratio (>106 for the positive scan and 104-106 for the negative scan) and long-term retention time indicated the high fidelity, low error, and high stability of the resistive memory devices. The memory behavior was attributed to an electric field-induced intramolecular charge transfer in an ordered donor-acceptor system, which provided the effective charge-transfer channels for injected charge carriers. This work represents the first example that explores the resistive memory properties of 2D PI-COF films, shedding light on the potential application of 2D COFs as information storage media.

In this paper, we proposed an unsupervised learning method for estimating the optical flow between video frames, especially to solve the occlusion problem. Occlusion is caused by the movement of an object or the movement of the camera, defined as when certain pixels are visible in one video frame but not in adjacent frames. Due to the lack of pixel correspondence between frames in the occluded area, incorrect photometric loss calculation can mislead the optical flow training process. In the video sequence, we found that the occlusion in the forward ($t\rightarrow t+1$) and backward ($t\rightarrow t-1$) frame pairs are usually complementary. That is, pixels that are occluded in subsequent frames are often not occluded in the previous frame and vice versa. Therefore, by using this complementarity, a new weighted loss is proposed to solve the occlusion problem. In addition, we calculate gradients in multiple directions to provide richer supervision information. Our method achieves competitive optical flow accuracy compared to the baseline and some supervised methods on KITTI 2012 and 2015 benchmarks. This source code has been released at https://github.com/jianfenglihg/UnOpticalFlow.git., 6 pages, 5 figures

The 2030 Agenda for Sustainable Development, adopted by the United Nations in 2015, set 17 global Sustainable Development Goals (SDGs) with 169 targets. 232 indicators were proposed as the global framework for monitoring the progress made in achieving SDGs. SDG targets and their corresponding indicators interact with each other forming an indivisible system. This study examined the interlinkages between SDG targets in China and used 51 indicators and relevant time-series data mapping with 108 SDG targets to quantify the network of interlinkages between SDG targets. Using Social Network Analysis and Principal Component Analysis, we identified 17 strategic targets and 17 core indicators. The core indicators can explain more than 95% of the variance of 51 indicators. Using core indicators can reduce statistical burden while satisfying the needs for SDG monitoring. Based on the dashboard indicating the synergies and trade-offs between 17 strategic targets and other targets, this paper recommended that China should set 17 strategic targets as priority areas and optimise the use of limited financial resources by maximising the synergies and minimising the trade-offs.

In this article, a series of tetra-functional quinoxaline-based benzoxazine/ monofunctional benzoxazine (P-a) blending resins were prepared. Differential scanning calorimeter (DSC), dynamic mechanical analyzer (DMA) and thermogravimetric analysis (TGA) were utilized to study the curing behavior of blends and heat resistance and thermal stability of the corresponding resins. The obtained blends show lower onset and peak exothermic temperatures than that of a phenol/phenylamine-based benzoxazine. Additionally, the cured blends possess much better heat resistance, higher char yields, and better thermal stability than monofunctional benzoxazine resins.

A new series of aniline and aniline-mixed tetrafunctional fluorene-based oxazine monomers were synthesized using 2,7-hydroxy-9,9-bis-(4-hydroxyphenyl) fluorene, paraformaldehyde, and primary amines (including aniline or aniline mixed with n-butylamine or n-octylamine composition). Fourier transform infrared spectroscopy and proton nuclear magnetic resonance spectroscopy were used to characterize the structure of the monomers. The copolymers were obtained by adding the monomers into a typical monofunctional polybenzoxazine (phenol-aniline-based benzoxazine). Differential scanning calorimetry, thermogravimetric analysis, and dynamic mechanical analysis were performed to study the thermal properties of the copolymers. The copolymers exhibited high glass transition temperature values (164–201°C). A good thermal stability was also obtained with a 5% weight loss temperature over 355°C and high char yields at 800°C (42–50%).

A series of tetra-functional fluorene-based benzoxazines containing both flexible linear aliphatic chain and rigid aromatic structure in their backbones were synthesized using mixed amines such as aniline and n-octylamine, 2,7-dihydroxy-9,9-bis-(4-hydroxyphenyl)fluorene (THPF) and paraformaldehyde as raw materials via Mannich reaction. The prepared benzoxazine monomers were identified by fourier transform infrared spectroscopy (FTIR), hydrogen nuclear magnetic resonance (1H NMR). In addition, the curing behavior, curing kinetics and rheological properties of these monomers as well as the thermal and mechanical properties of their cured resins were studied using a rheometer, differential scanning calorimeter (DSC), thermogravimetric analyzer (TGA), and dynamic thermomechanical analyzer (DMA). The newly developed benzoxazines show good processibility, excellent thermal stability and high glass transition temperature (Tg) values ranging from 291 to 307 °C. By varying the proportion of n-octylamine to aniline, the properties of these resins are tuned.

This study deals with the influences of both the length of the aliphatic spacer within the phthalonitrile monomers backbones, and the amount of the silane surface modified zirconia nanoparticles on the mechanical and thermal properties of the so-called second generation phthalonitrile resins. Investigation on the curing behavior under differential scanning calorimeter outlined an important gain in the processability as the aliphatic spacer became longer. Results from the mechanical tests revealed that changing the length of the aliphatic spacer affects the mechanical properties in different ways. For instance, as the aliphatic spacer became longer, the toughness state was enhanced. At the same time, the tensile modulus and stress as well as the microhardness values were slightly reduced. It was also noticed that the introduction of the reinforcing phase caused an increase in all the tested mechanical properties. Furthermore, results from the thermogravimetric analysis and dynamic mechanical analysis revealed that reducing the length of the aliphatic spacer and adding nanofillers caused an increase in the thermal stability, storage modulus, and glass transition temperature. Moreover, a morphological study has been conducted under scanning electron microscope and transmission electron microscope to put in light the mechanisms of enhancements. Finally, this study demonstrated that the excellent properties of the phthalonitrile resins can be tailored by two ways either by monomers design or by inorganic nanoparticles reinforcement.

Matching suitability analysis is a key issue of INS/SAR integrated navigation mode. The existing suitability area selection methods use the simulated real-time image to calculate the matching probability of the scene area and further label it “suitability” or “unsuitability”. If the imaging mode of the simulated image is the same as that of the real image, the suitability area selection model based on the simulated real-time image works well. Otherwise, the model is impractical. In order to address this issue, a novel method is proposed in this paper. The sample dataset is built on the actual flight real-time images, and a hybrid feature selection method based on D-Score and SVM is used to select the suitability features and build the suitability area selection model simultaneously. Experimental results show that the consistency between the prediction results of the model and the ones experts label reaches 81.92%.

A new kind of thermally-conductive polymer nanocomposite was produced by reinforcing the typical polybenzoxazine matrix P(BA-a) with silane-treated Si 3 N 4 nanoparticles using a compression molding technique. By adding 0 to 70% nano-Si 3 N 4 , the nanocomposite׳s thermal conductivity was enhanced from 0.18 to 5.78 W/m K. Based on the analyzed morphology, the Si 3 N 4 nanoparticles tend to form continuous conductive chains within the polybenzoxazine matrix, which are responsible for the improved thermal conductivity of the polybenzoxazine resin. Also, these new nanocomposites exhibited improved dielectric, coefficient of thermal expansion, and flexural properties. The thermal stability of these materials significantly improved upon the addition of the nano-Si 3 N 4 fillers. Their unique thermal and dielectric properties make them suitable for microelectronic packaging applications.

A new type of composite based on phthalonitrile resin reinforced with silicon carbide (SiC) microparticles was prepared. For various weight ratios ranging between 0% and 20%, the effect of the micro-SiC particles on the mechanical and thermal properties has been studied. Results from thermal analysis revealed that the starting decomposition temperature and the residual weight were significantly improved upon adding the reinforcing phase. At the maximum micro-SiC loading, dynamic mechanical analysis (DMA) showed an important enhancement in both the storage modulus and glass transition temperature ( T g ), reaching 3.1 GPa and 338 °C, respectively. The flexural strength and modulus as well as the microhardness were significantly enhanced by adding the microfillers. Tensile test revealed enhancements in the composites toughness upon adding the microparticles. Polarization optical microscope (POM) and scanning electron microscope (SEM) analysis confirmed that mechanical and thermal properties improvements are essentially attributed to the good dispersion and adhesion between the particles and the resin.

A series of fluorene-based benzoxazine copolymers were synthesized from the mixture of 9,9-bis(4-hydroxyphenyl)fluorene and bisphenol A, and 4,4′-diaminodiphenyloxide and paraformaldehyde. And the cured polybenzoxazine films derived from these copolymers were also obtained. Fourier transform infrared spectroscopy (FTIR) and hydrogen nuclear magnetic resonances confirmed the structure of these benzoxazines. Their molecular weight was estimated by gel permeation chromatography. The curing behavior of the precursors was monitored by FTIR and differential scanning calorimetry. Dynamic mechanical analysis and thermogravimetric analysis were performed to study the thermal properties of the cured polymers. The cured polybenzoxazines exhibit excellent heat resistance with glass transition temperatures (Tg) of 286–317°C, good thermal stability along with the values of 5% weight loss temperatures (T5) over 340°C, and high char yield over 50% at 800°C. The mechanical properties of the cured polymers were also measured by bending tests. Copyright © 2015 John Wiley & Sons, Ltd.

A novel tetraphenol fluorene, 2,7-dihydroxy-9,9-bis-(4-hydroxyphenyl)fluorene (THPF), was synthesized via the condensation reaction of 2,7-dihydroxy-9-fluorenone and phenol in the presence of a strong acidic cation exchange resin and 3-mercaptopropionic acid. Thus, a novel tetrafunctional oxazine monomer containing benzoxazine and fluorene-oxazine (t-BF-b) was prepared for the first time using a Mannich condensation reaction of THPF with paraformaldehyde and n-butylamine. The chemical structures of THPF and t-BF-b were characterized by Fourier transform infrared (FTIR) spectroscopy, elemental analysis, 1H and 13C nuclear magnetic resonance (NMR). The viscosity–temperature properties and the polymerization behavior of t-BF-b as well as the thermal and mechanical properties of its cured polymer (poly(t-BF-b)) were studied by rheometry, FTIR, 1H NMR, differential scanning calorimetry (DSC), thermogravimetric analysis (TGA), and dynamic mechanical analysis (DMA). The results show that poly(t-BF-b) displays a lower melting point and wider processing window. The oxazine rings of fluorene-oxazine possess higher reactivity and lower polymerization temperature than those of benzoxazine. Also, its cured poly(t-BF-b) exhibits a higher glass transition temperature than its corresponding bifunctional polybenzoxazines without sacrificing any thermal properties in spite of the introduction of more flexible aliphatic groups into polymer chains.

A series of fluorene-based benzoxazine-endcapped copoly(ether ketone ketone)s (co-PEKKs) were successfully synthesized, in which the co-PEKKs were prepared with different molar ratios of 2,7-dihydroxy-9-fluorenone to bisphenol-A. The obtained phenol-ended co-PEKKs reacted with formaldehyde and n-butylamine to yield the benzoxazine-endcapped co-PEKKs (B-PEKKs). The structure of these oligomers was studied by Fourier transform infrared spectroscopy and 1H nuclear magnetic resonance spectroscopy, and the molecular mass of these phenol-ended co-PEKKs was detected by gel permeation chromatography. The thermal behaviors of these B-PEKKs were determined by differential scanning calorimetry. In addition, the thermal and dynamic mechanical properties of the polymer films were investigated by thermogravimetric analysis and dynamic mechanical analysis. The results indicate that the incorporation of co-PEKKs within the benzoxazine backbone considerably improves the toughness of the fluorene-containing polybenzoxazines without much sacrifice of their thermal properties. Furthermore, as the content of 2,7-dihydroxy-9-fluorenone increases, the T5 values decrease, while the Tg values and char yields of the prepared polymers increase to reach their highest values at 256 °C and 64 %, respectively.

Silicon nitride (SN) nanoparticle-reinforced polybenzoxazine nanocomposites were produced with good processability at various fractions of nano-SN ranging between 0 and 30 wt% via a compression molding technique. At the maximum nano-SN loading, the DMA of the nanocomposites showed an increase of 2 GPa and 47 °C in stiffness and glass transition temperature (Tg), respectively. In addition, increasing the nano-SN fraction to 30 wt% significantly enhanced both the tensile strength and the micro-hardness of the nanocomposites. The TGA results revealed that the thermal stability of these nanocomposites was highly improved upon the addition of the reinforcing phase. The enhancements in the thermal and mechanical properties of these nanocomposites are attributed to the uniform nanoparticle dispersion and to the strong matrix–particle interactions based on the SEM and TEM results. Therefore, these nanocomposites are more suitable for high performance coating applications.

The recent development of high-molecular weight benzoxazine precursors is discussed in this chapter. Also, the studies of fluorene-based benzoxazine monomers and main chain, side chain, and telechelic high molecular weight benzoxazine oligomers and co-oligomers containing fluorene groups are reviewed in detail. Compared with fluorene-based benzoxazine and fluorene-oxazine monomers, the fluorene-containing high-molecular weight benzoxazine precursors exhibit better thermal properties. Then these benzoxazine precursors are incorporated into traditional monofunctional benzoxazine, P-a. The curing behavior of the obtained blends are studied by differential scanning calorimeter (DSC). The thermal and mechanical properties of the cured blends are evaluated by thermogravimetric analysis (TGA) and dynamic mechanical analysis (DMA). These produced blend resins could be good candidates for high performance thermosets.

With the increasing environmental awareness and the fast depletion of petroleum resources, bio-based polymers synthesized from renewably derived chemicals have entered our awareness gradually. Among all renewable resources, furan-containing compounds such as furfuryl alcohol and furfurylamine, which is derived from a variety of agricultural by-products such as corncobs and wheat bran, have been widely used as the precursors of polymers with high-performance. In this chapter, the structure, polymerization, thermal, and thermomechanical properties of various furan-containing, bio-based benzoxazines are briefly introduced. The synthesis, characterization, polymerization behavior, thermal, and thermomechanical properties of a series of novel furan-containing, tetrafunctional fluorene-based oxazine monomers and their polymers that simultaneously contain benzoxazine and fluorene-oxazine are mainly described. By means of the introduction of furan rings and more oxazine rings, these cured polymers and copolymers with typical monofunctional polybenzoxazine (P-a) display an extremely higher Tg value (442°C) and a much better thermal stability than those of traditional difunctional and multifunctional polybenzoxazines due to much more cross-linked sites.

For various molecular ratios ranging between 0% and 30%, the effect of Crab Shell Particles (CSP) on the thermal and thermomechanical properties of polybenzoxazine has been studied. The high contents of CaCO3 particles in CSP clearly improve both the thermomechanical and thermal properties of the polybenzoxazine corresponding to a gain in storage modulus and glass transition temperature of 137% and 25%, respectively, at 30% CSP loading. Thermal analysis reveals that the polybenzoxazine matrix filled with CSP also increases the initial decomposition temperatures and the char yield of composites which reaches 49% at maximum CSP content.

This study investigates the effects of chitosan particles (CTS) on the curing behavior, thermal and thermomechanical properties of polybenzoxazine matrix. The morphological, thermomechanical, and thermal properties of the blends are analyzed using scanning electron microscopy (SEM), dynamic scanning calorimetry (DSC), dynamical mechanical analyzer (DMA), and thermogravimetric analysis (TGA). The results show that the −NH2 groups on the chitosan can act as an active crosslinking position and hydrogen bonding. The SEM micrographs reveal good compatibility between the blend components. Furthermore, the values of glass transition temperatures, char yields, and storage moduli of the cured blends are found to be increased with the increase of CTS contents to reach 191℃, 34%, 4.3 GPa, respectively, at 10% of CTS content.

A novel furan-containing tetrafunctional fluorene-based benzoxazine monomer with bisphenol- and diamine-type oxazine rings was successfully prepared. The resulting polybenzoxazine exhibits extremely higher glass transition temperature (440 °C) and better thermal stability than difunctional furan-containing fluorene-based and traditional multifunctional benzoxazine resins.

The purpose of this study is to develop a multi-physiological parameters monitoring system using wireless body area network (WBAN), which is aided to assess motor functional rehabilitation process in robot-aided rehabilitation training. Firstly, electrocardiogram (ECG), pulse, skin resistance, and body-temperature physiological parameters monitoring hardware modules are designed. Then, system software including WBAN nodes communication and monitoring software are programmed. Finally, experiments based on a healthy subject are conducted and the discussed multi-physiological parameters are real-time acquired and analyzed. Experimental results verify the effectiveness of the proposed monitoring system, and the measured multi-physiological parameters can be potentially used to remotely and continuously monitor and assess of stroke rehabilitation process.