Deep learning has recently attracted much attention due to its excellent performance in processing audio, image, and video data. However, few studies are devoted to the field of automatic modulation classification (AMC). It is one of the most well-known research topics in communication signal recognition and remains challenging for traditional methods due to complex disturbance from other sources. This paper proposes a heterogeneous deep model fusion (HDMF) method to solve the problem in a unified framework. The contributions include the following: (1) a convolutional neural network (CNN) and long short-term memory (LSTM) are combined by two different ways without prior knowledge involved; (2) a large database, including eleven types of single-carrier modulation signals with various noises as well as a fading channel, is collected with various signal-to-noise ratios (SNRs) based on a real geographical environment; and (3) experimental results demonstrate that HDMF is very capable of coping with the AMC problem, and achieves much better performance when compared with the independent network.

Abstract:The semantic segmentation of remote sensing images faces two major challenges: high inter-class similarity and interference from ubiquitous shadows. In order to address these issues, we develop a novel edge loss reinforced semantic segmentation network (ERN) that leverages the spatial boundary context to reduce the semantic ambiguity. The main contributions of this paper are as follows: (1) we propose a novel end-to-end semantic segmentation network for remote sensing, which involves multiple weighted edge supervisions to retain spatial boundary information; (2) the main representations of the network are shared between the edge loss reinforced structures and semantic segmentation, which means that the ERN simultaneously achieves semantic segmentation and edge detection without significantly increasing the model complexity; and (3) we explore and discuss different ERN schemes to guide the design of future networks. Extensive experimental results on two remote sensing datasets demonstrate the effectiveness of our approach both in quantitative and qualitative evaluation. Specifically, the semantic segmentation performance in shadow-affected regions is significantly improved.

In recent years, it is heartening to witness that carbon quantum dots (CQDs), a rising star in the family of carbon nanomaterials, have displayed tremendous applications in bioimaging, biosensing, drug delivery, optoelectronics, photovoltaics and photocatalysis. However, the investigations toward self-assembly of CQDs are still in their infancy. The participation of CQDs can bring additional functions to supramolecular self-assemblies, with photoluminescent property as the most exciting aspect. Here, we introduce CQDs into two types of classic colloidal systems containing low molecular weight surfactant and gelator to construct fluorescent vesicles and chiral hydrogels. The CQD-based vesicles were constructed through electrostatic interaction between the positively charged CQDs with peripherally substituted imidazolium cations and a negatively-charged biosurfactant, i.e., sodium deoxycholate (NaDC). The chiral hydrogels were prepared by increasing the concentration of NaDC and addition of a tripeptide (glutathione, GSH). It was found that both the hydrogels and corresponding xerogels are highly photoluminescent. A solid sensing system was prepared by coating a uniform layer of the hydrogel onto the silica gel plates by doctor blade technique followed by air-drying, which was then utilized to semiquantitatively detect Cu2+ in aqueous solutions.

Conventionally, amphiphiles are composed of hydrophobic and hydrophilic units. They are able to exhibit a wide variety of structures depending on the environment. Such features have been applied in supramolecular chemistry, by which apolar and polar groups are implemented in the molecular design. Here we present an attractive approach to introduce unique amphiphilicity. Relatively simple fullerene (C(60)) derivatives that bear long aliphatic chains behave as uncommon surfactants in organic media. Although two hydrophobic units are used to assemble the derivatives, slight differences in their polarity and chemical nature may make them incompatible and thus arrange them in microphase-separated mesostructures as lamellar ones. These assemblies are maintained by relatively weak forces, pi-pi interactions among C(60) moieties and van der Waals forces between alkyl chains. Therefore, the derivatives can undergo "supramolecular polymorphism" by which different supramolecular assemblies arise by changing the conditions of assembly. A simple modification in their substituent motif of derivatives influences the intermolecular interactions and provides a wide variety of supramolecular materials.

Mixing negatively charged carbon quantum dots with a zwitterionic surfactant in water produces a variety of supramolecular structures, which are photoluminescent and show a reversible response to pH.

Robotic vision-based crack detection in concrete bridges is an essential task to preserve these assets and their safety. The conventional human visual inspection method is time consuming and cost inefficient. In this paper, we propose a robust algorithm to detect cracks in a pixel-wise manner from real concrete surface images. In practice, crack detection remains challenging in the following aspects: (1) detection performance is disturbed by noises and clutters of environment; and (2) the requirement of high pixel-wise accuracy is difficult to obtain. To address these limitations, three steps are considered in the proposed scheme. First, a local pattern predictor (LPP) is constructed using convolutional neural networks (CNN), which can extract discriminative features of images. Second, each pixel is efficiently classified into crack categories or non-crack categories by LPP, using as context a patch centered on the pixel. Lastly, the output of CNN—i.e., confidence map—is post-processed to obtain the crack areas. We evaluate the proposed algorithm on samples captured from several concrete bridges. The experimental results demonstrate the good performance of the proposed method.

A new series of N-methylfulleropyrrolidines bearing oligo(poly(ethylene oxide))-appended Percec monodendrons (fulleromonodendrons, 4a-f) have been synthesized. The substituted position of the oligo(poly(ethylene oxide)) chain(s) on the phenyl group of the Percec monodendron for 4a-f was varied, which is at the 4-, 2,4-, 3,5-, 3,4,5-, 2,3,4- and 2,4,6- position, respectively. 4a-e are obtained as solids at 25 °C and can self-organize into lamellar phases as revealed by X-ray diffraction (XRD) and small-angle X-ray scattering (SAXS) measurements, while 4f appears as a viscous liquid. The substitution patterns of the oligo(poly(ethylene oxide)) chain(s) also significantly influence the solubility of 4a-f, especially in ethanol and water. Formation of self-organized supramolecular structures of 4d and 4e in water as well as 4d in ethanol is evidenced from UV-vis and dynamic light scattering (DLS) measurements. Further studies in water using various imaging techniques including transmission electron microscopy (TEM), freeze-fracture TEM (FF-TEM), cryo-TEM and atomic force microscopy (AFM) observations revealed the formation of well-defined vesicles for 4d and plate-like aggregates for 4e, indicating that the aggregation behavior of the fulleromonodendrons is highly dependent on their molecular structures. For 4d in ethanol, only irregular aggregates were noticed, indicating the solvent also plays a role on regulating the aggregation behavior. After functionalization with the Percec monodendrons, 4a-f can preserve the intriguing electrochemical properties of pristine C60 as revealed by cyclic voltammetries. The thermotropic properties of 4a-f have also been investigated. It was found that all of them show good thermal stability, but no mesophases were detected within the investigated temperature ranges.

No abstract