SFA-Net: A Selective Features Absorption Network for Object Detection in Rainy Weather Conditions.

In recent years, object detection approaches using deep convolutional neural networks (CNNs) have derived major advances in normal images. However, such success is hardly achieved with rainy images due to lack of visibility. Aiming to bridge this gap, in this article, we present a novel selective features absorption network (SFA-Net) to improve the performance of object detection not only in rainy weather conditions but also in favorable weather conditions. SFA-Net accomplishes this objective by utilizing three subnetworks, where the feature selection subnetwork is concatenated with the object detection subnetwork through the feature absorption subnetwork to form a unified model. To promote further advancement in object detection impaired by rain, we propose a large-scale rainy image dataset, named srRain, which contains both synthetic rainy images and real-world rainy images for training and testing purposes. srRain is comprised of 25 900 rainy images depicting diverse driving scenarios in the presence of rain with a total of 181 164 instances interpreting five common object categories. Experimental results display that our SFA-Net reaches the highest mean average precision (mAP) of 77.53% on a normal image set, 62.52% on a synthetic rainy image set, 37.34% on a collected natural rainy image set, and 32.86% on a published real rainy image set, surpassing current state-of-the-art object detectors and the combination of image deraining and object detection models while retaining a high speed.

Genotype-phenotype characteristics of Vietnamese patients diagnosed with Charcot-Marie-Tooth disease.

BACKGROUND: Charcot-Marie-Tooth (CMT) disease is one of the most common hereditary neuropathies. Identifying causative mutations in CMT is essential as it provides important information for genetic diagnosis and counseling. However, genetic information of Vietnamese patients diagnosed with CMT is currently not available. METHODS: In this study, we described the clinical profile and determined the mutation spectrum of CMT in a cohort of Vietnamese patients with CMT by using a combination of multiplex ligation-dependent probe amplification and next-generation sequencing targeting 11 genes PMP22, MPZ, EGR2, NEFL, MFN2, GDAP1, GARS, MTMR2, GJB1, RAB7A, LITAF. RESULTS: In 31 CMT cases, the mutation detection rate was 42% and the most common genetic aberration was PMP22 duplication. The pedigree analysis showed two de novo mutations c.64C > A (p.P22T) and c.281delG (p.G94Afs*17) in the NEFL and PMP22 genes, respectively. CONCLUSION: The results of this study once again emphasize the important role of molecular diagnosis and provide preliminary genetic data on Vietnamese patients with CMT.

Deeply Cyclable and Ultrahigh-Rate Lithium Metal Anodes Enabled by Coaxial Nanochamber Heterojunction on Carbon Nanofibers.

Lithium metal anodes (LMAs) are the most promising candidates for high-energy-density batteries due to the high theoretical specific capacity and lowest potential. However, the practical application of LMAs is hampered by the short lifespan and unsatisfactory lithium utilization (<50%). An oxide-oxide heterojunction enhanced with nanochamber structure design is proposed to improve lithium utilization and cycling performance of LMA under ultrahigh rates. Typically, a MnO2 -ZnO heterojunction provides high binding energy for strong absorption of Li-ions and intimately bonded interfaces for fast transfer of electrons. Under the guidance of the smooth Li-ion migration and rapid electron flow, the Li metal can be restricted as thin layers within submicro scale in nanochambers with constrain boundary and stress dissipation, inhibiting the local agglomeration and blocking. Thus, the lithiophilic active sites can be effectively exposed to the Li-ions within submicro scale, improving the reversible conversion for high lithium utilization during long-term cycling. As such, the Li@MnZnO/CNF electrode achieves a high lithium utilization of 70% at a record-high current density of 50 mA cm-2 with areal capacity of 10 mAh cm-2 . This work offers an avenue to improve lithium utilization for long-lifespan LMAs working under high current densities and capacities.

An Advanced Noise Reduction and Edge Enhancement Algorithm.

Complementary metal-oxide-semiconductor (CMOS) image sensors can cause noise in images collected or transmitted in unfavorable environments, especially low-illumination scenarios. Numerous approaches have been developed to solve the problem of image noise removal. However, producing natural and high-quality denoised images remains a crucial challenge. To meet this challenge, we introduce a novel approach for image denoising with the following three main contributions. First, we devise a deep image prior-based module that can produce a noise-reduced image as well as a contrast-enhanced denoised one from a noisy input image. Second, the produced images are passed through a proposed image fusion (IF) module based on Laplacian pyramid decomposition to combine them and prevent noise amplification and color shift. Finally, we introduce a progressive refinement (PR) module, which adopts the summed-area tables to take advantage of spatially correlated information for edge and image quality enhancement. Qualitative and quantitative evaluations demonstrate the efficiency, superiority, and robustness of our proposed method.

A Fishing-Net-Like 3D Host for Robust and Ultrahigh-Rate Lithium Metal Anodes.

Constructing an architectural host is demonstrated to be an effective strategy for long-life lithium metal anodes (LMAs). Herein, an integrated 3D host for stable and ultrahigh-rate LMAs is developed by a binary highly conductive network of 2D reduced graphene oxide (rGO) and 1D carbon nanofibers (CNF) anchored with 0D ultrasmall MgZnO nanoparticles (MgZnO/CNF-rGO). A facile net-fishing strategy is proposed to combine the rGO nanosheets with free-standing CNF matrix as interconnected paths for fast electron transport. Notably, serving as Li nucleation sites, the superlithiophilic MgZnO nanoparticles are uniformly distributed and tightly contacted with the conductive matrix without agglomeration due to the rGO confinement. Such a delicate nanoscale combination guarantees the effective transportation and uniform deposition of Li-ions in the inner surface of the host. The symmetric cell of Li@MgZnO/CNF-rGO exhibits a long lifespan above 1450 cycles under an ultrahigh current density of 50 mA cm-2 with an areal capacity of 1.0 mAh cm-2 . Impressively, it also delivers a high reversible capacity of 10 mAh cm-2 at 50 mA cm-2 . This work offers an avenue to promise the prospect for practical LMAs working under high rates and capacities.

DSNet: Joint Semantic Learning for Object Detection in Inclement Weather Conditions.

In the past half of the decade, object detection approaches based on the convolutional neural network have been widely studied and successfully applied in many computer vision applications. However, detecting objects in inclement weather conditions remains a major challenge because of poor visibility. In this article, we address the object detection problem in the presence of fog by introducing a novel dual-subnet network (DSNet) that can be trained end-to-end and jointly learn three tasks: visibility enhancement, object classification, and object localization. DSNet attains complete performance improvement by including two subnetworks: detection subnet and restoration subnet. We employ RetinaNet as a backbone network (also called detection subnet), which is responsible for learning to classify and locate objects. The restoration subnet is designed by sharing feature extraction layers with the detection subnet and adopting a feature recovery (FR) module for visibility enhancement. Experimental results show that our DSNet achieved 50.84 percent mean average precision (mAP) on a synthetic foggy dataset that we composed and 41.91 percent mAP on a public natural foggy dataset (Foggy Driving dataset), outperforming many state-of-the-art object detectors and combination models between dehazing and detection methods while maintaining a high speed.

A Triple-Gradient Host for Long Cycling Lithium Metal Anodes at Ultrahigh Current Density.

The viable Li metal anodes (LMAs) are still hampered by the safety concerns resulting from fast Li dendrite growth and huge volume expansion during cycling. Herein, carbon nanofiber matrix anchored with MgZnO nanoparticles (MgZnO/CNF) is developed as a flexible triple-gradient host for long cycling LMAs. The superlithiophilic MgZnO nanoparticles significantly increase the wettability of CNF for fast and homogeneous infusion with molten Li. The in-built potential and lithiophilic gradients constructed after an in situ lithiation of MgZnO and CNF enable nearly zero Li nucleation overpotential and homogeneous deposition of lithium at different scales. As such, the LMAs based on MgZnO/CNF achieve long cycling life and small overpotential even at a record-high current density of 50 mA cm-2 and a high areal capacity of 10 mAh cm-2 . A full cell paring with this designed LMA and LiFePO4 exhibits a capacity retention up to 82% after 600 cycles at a high rate of 5 C. A Li-ion capacitor also shows an impressive capacity retention of 84% at 5 A g-1 after 10 000 cycles. Such a Li@MgZnO/CNF anode is a promising candidate for Li-metal energy storage systems, especially working under ultrahigh current density.

Novel GDAP1 Mutation in a Vietnamese Family with Charcot-Marie-Tooth Disease.

BACKGROUND: Mutations of GDAP1 gene cause autosomal dominant and autosomal recessive Charcot-Marie-Tooth (CMT) disease and over 80 different mutations have been identified so far. This study analyzed the clinical and genetic characteristics of a Vietnamese CMT family that was affected by a novel GDAP1 mutation. METHODS: We present three children of a family with progressive weakness, mild sensory loss, and absent tendon reflexes. Electrodiagnostic analyses displayed an axonal type of neuropathy in affected patients. Sequencing of GDAP1 gene was requested for all members of the family. RESULTS: All affected individuals manifested identical clinical symptoms of motor and sensory impairments within the first three years of life, and nerve conduction study indicated the axonal degeneration. A homozygous GDAP1 variant (c.667_671dup) was found in the three affected children as recessive inheritance pattern. The mutation leads to a premature termination codon that shortens GDAP1 protein (p.Gln224Hisfs∗37). Further testing showed heterozygous c.667_671dup variant in the parents. DISCUSSION: Our study expands the mutational spectrum of GDAP1-related CMT disease with the new and unreported GDAP1 variant. Alterations in GDAP1 gene should be evaluated as CMT causing variants in the Vietnamese population, predominantly axonal form of neuropathy in CMT disease.

Effects of Electrospun Carbon Nanofibers' Interlayers on High-Performance Lithium-Sulfur Batteries.

Two different interlayers were introduced in lithium-sulfur batteries to improve the cycling stability with sulfur loading as high as 80% of total mass of cathode. Melamine was recommended as a nitrogen-rich (N-rich) amine component to synthesize a modified polyacrylic acid (MPAA). The electrospun MPAA was carbonized into N-rich carbon nanofibers, which were used as cathode interlayers, while carbon nanofibers from PAA without melamine was used as an anode interlayer. At the rate of 0.1 C, the initial discharge capacity with two interlayers was 983 mAh g-1, and faded down to 651 mAh g-1 after 100 cycles with the coulombic efficiency of 95.4%. At the rate of 1 C, the discharge capacity was kept to 380 mAh g-1 after 600 cycles with a coulombic efficiency of 98.8%. It apparently demonstrated that the cathode interlayer is extremely effective at shutting down the migration of polysulfide ions. The anode interlayer induced the lithium ions to form uniform lithium metal deposits confined on the fiber surface and in the bulk to strengthen the cycling stability of the lithium metal anode.

In-situ growth of MnO2 crystals under nanopore-constraint in carbon nanofibers and their electrochemical performance.

Growing MnO2 nanocrystals in the bulk of porous carbon nanofibers is conducted in a KMnO4 aqueous solution aimed to enhance the electrochemical performance of MnO2. The rate of redox reaction between KMnO4 and carbon was controlled by the concentration of KMnO4 in a neutral solution. The MnO2 nanoparticles grow along with (211) crystal faces when the redox reaction happens on the surface of fibers under 1D constraint, while the nanoparticles grow along with (200) crystal faces when the redox reaction happens in the bulk of fibers under 3D constraint. The composite, where MnO2 nanoparticles are formed in the bulk under a constraint, yields an electrode material for supercapacitors showing good electron transport, rapid ion penetration, fast and reversible Faradaic reaction, and excellent rate performance. The capacitance of the composite electrode could be 1282 F g-1 under a current density of 0.2 A g-1 in 1 M Na2SO4 electrolyte. A symmetric supercapacitor delivers energy density of 36 Wh kg-1 with power density of 39 W kg-1, and can maintain 7.5 Wh kg-1 at 10.3 kW kg-1. It exhibits an excellent electrochemical cycling stability with 101% initial capacitance and 95% columbic efficiency even after 1000 cycles of charge/discharge.