Unsupervised Sentence Representation Learning with Frequency-induced Adversarial tuning and Incomplete sentence filtering.

Pre-trained Language Model (PLM) is nowadays the mainstay of Unsupervised Sentence Representation Learning (USRL). However, PLMs are sensitive to the frequency information of words from their pre-training corpora, resulting in anisotropic embedding space, where the embeddings of high-frequency words are clustered but those of low-frequency words disperse sparsely. This anisotropic phenomenon results in two problems of similarity bias and information bias, lowering the quality of sentence embeddings. To solve the problems, we fine-tune PLMs by leveraging the frequency information of words and propose a novel USRL framework, namely Sentence Representation Learning with Frequency-induced Adversarial tuning and Incomplete sentence filtering (Slt-fai). We calculate the word frequencies over the pre-training corpora of PLMs and assign words thresholding frequency labels. With them, (1) we incorporate a similarity discriminator used to distinguish the embeddings of high-frequency and low-frequency words, and adversarially tune the PLM with it, enabling to achieve uniformly frequency-invariant embedding space; and (2) we propose a novel incomplete sentence detection task, where we incorporate an information discriminator to distinguish the embeddings of original sentences and incomplete sentences by randomly masking several low-frequency words, enabling to emphasize the more informative low-frequency words. Our Slt-fai is a flexible and plug-and-play framework, and it can be integrated with existing USRL techniques. We evaluate Slt-fai with various backbones on benchmark datasets. Empirical results indicate that Slt-fai can be superior to the existing USRL baselines.

Pillar[5]arene-segregated ion pairs for enhanced cycloaddition of epoxides with CO2.

A supramolecular approach using a polyviologen-pillar[5]arene complex as segregated ion pairs was shown to be highly efficient for the conversion of CO2 with epoxides into cyclic carbonates without the need for metals or solvents. The enhanced catalytic performance was achieved by cooperative ion pair segregation and CO2 fixation.

Fault Detection in Power Distribution Networks Based on Comprehensive-YOLOv5.

Real-time fault detection in power distribution networks has become a popular issue in current power systems. However, the low power and computational capabilities of edge devices often fail to meet the requirements of real-time detection. To overcome these challenges, this paper proposes a lightweight algorithm, named Comprehensive-YOLOv5, for identifying defects in distribution networks. The proposed method focuses on achieving rapid localization and accurate identification of three common defects: insulator without loop, cable detachment from the insulator, and cable detachment from the spacer. Based on the You Only Look Once version 5 (YOLOv5) algorithm, this paper adopts GhostNet to reconstruct the original backbone of YOLOv5; introduces Bidirectional Feature Pyramid Network (BiFPN) structure to replace Path Aggregation Network (PANet) for feature fusion, which enhances the feature fusion ability; and replaces Generalized Intersection over Union GIOU with Focal Extended Intersection over Union (Focal-EIOU) to optimize the loss function, which improves the mean average precision and speed of the algorithm. The effectiveness of the improved Comprehensive-YOLOv5 algorithm is verified through a "morphological experiment", while an "algorithm comparison experiment" confirms its superiority over other algorithms. Compared with the original YOLOv5, the Comprehensive-YOLOv5 algorithm improves mean average precision (mAP) from 88.3% to 90.1% and increases Frames per second (FPS) from 20 to 52 frames. This improvement significantly reduces false positives and false negatives in defect detection. Consequently, the proposed algorithm enhances detection speed and improves inspection efficiency, providing a viable solution for real-time detection and deployment at the edge of power distribution networks.

Focus stacking in non-contact dermoscopy.

Dermoscopy is the main tool for early detection of skin cancer. Non-contact dermoscopes often suffer from a small depth of field leading to images of skin topographies with regions that are not in focus. We aim to provide an easy-to-implement focus stacking-based approach to ensure all-in-focus images from a non-contact dermoscope. Further, we aim to extract additional information about the skin topography from the image stacks. The focus stacking procedure itself is implemented in a non-contact dermoscope with an electrically adjustable focus realized by using a tunable liquid lens. We show that all-in-focus imaging is possible for non-contact dermoscopy and deliver a method to extract topographical information for dermatologists from the acquired image stacks. Our finding indicate that the approach can be valuable for non-contact dermoscopic examination as well as for the early detection of skin diseases such as cancer as it possible to derive hyperfocus images and information on the skin topography. With this, we were able to develop a software for the acquisition of the raw image data and its processing into a high resolution hyperresolution dermoscopic image. In the next steps, we plan to apply the approach in the clinical environment for skin cancer diagnostics or imaging of inflammatory skin diseases.

Nitrogen addition enhanced the polycyclic aromatic hydrocarbons dissipation through increasing the abundance of related degrading genes in the soils.

High concentrations of Polycyclic Aromatic Hydrocarbons (PAHs) in the soils cause significant threats to human health. Since nitrogen plays a crucial role in controlling microbial composition and functions in terrestrial ecosystems, bio-stimulation based on nitrogen has been used in PAHs contaminated environments remediation. Recent studies show that microbial community composition and organic pollutants dissipation correlate with nitrogen addition. Here, we investigated the effect of nitrogen addition on the abundance of microbial community, degrading genes, and their relationship to PAHs dissipation. After a 32-day experiment, PAHs residues in nitrogen treatment soil were reduced by 23.23%-34.21%. The application of 80 mg·kg-1 nitrate and ammonium nitrogen resulted in higher PAHs removal efficiency, and the dissipation rate of PAHs was 59.61% and 62.09%, respectively. Nitrogen application could improve the abundance and the diversity of soil microbial community. Degrading genes involved in PAH detoxification were enhanced after nitrogen addition, particularly those encoding ring-hydroxylating and catechol dioxygenases such as nahAc and nidA, thus, accelerating PAH dissipation in the soil. The results will facilitate the development of beneficial microbiome-based remediation strategies and improve agricultural production safety in PAHs-contaminated soils.

The impact of different IPV-OPV sequential immunization programs on hepatitis A and hepatitis B vaccine efficacy.

In recent years, the Global Polio Eradication Initiative has gradually implemented a global shift in polio immunization programs. Few studies cover polio immunization program impacts on the efficacy of other vaccines. This study investigated whether polio immunization programs affected hepatitis A (HepA) and hepatitis B (HepB) vaccination efficacy. Serum samples were collected from 968 infants before the first dose of polio vaccine, 28 days after completing primary polio immunization, and at 24 months old. Infants were classified into six polio immunization program groups: 1sIPV+2bOPV, 2sIPV+1bOPV, 2sIPV+1tOPV, 1cIPV+2bOPV, 2cIPV+1bOPV, and 2cIPV+1tOPV (sIPV: Sabin inactivated poliovirus vaccine; cIPV: Salk inactivated poliovirus vaccine; b, bivalent; t, trivalent; OPV, oral polio vaccine). No significant differences existed in antibody titers against HepA virus (anti-HAV) among the polio immunization program groups at any of the three time points (pre-first dose [p = 0.412], 28 days after primary immunization [p = 0.676], 24 months old [p = 0.556]). Before the first dose (p = 0.178) and at age 24 months (p = 0.987), no significant differences existed in HepB surface antibody (HBsAb) titers between the six polio immunization program groups). Twenty-eight days after primary immunization, no significant difference existed in HBsAb titers between groups after Bonferroni correction. Following HepA and HepB immunization, anti-HAV and HBsAb positivity reached > 98% in all groups, reflecting effective immunization. Our data suggest that different polio immunization programs did not affect HepA and HepB vaccine efficacy; HepA and HepB vaccines maintained high effectiveness irrespective of polio immunization program. This trial was registered on Clinical Trials.gov: NCT03614702.

Understanding building-occupant-microbiome interactions toward healthy built environments: A review.

Built environments, occupants, and microbiomes constitute a system of ecosystems with extensive interactions that impact one another. Understanding the interactions between these systems is essential to develop strategies for effective management of the built environment and its inhabitants to enhance public health and well-being. Numerous studies have been conducted to characterize the microbiomes of the built environment. This review summarizes current progress in understanding the interactions between attributes of built environments and occupant behaviors that shape the structure and dynamics of indoor microbial communities. In addition, this review also discusses the challenges and future research needs in the field of microbiomes of the built environment that necessitate research beyond the basic characterization of microbiomes in order to gain an understanding of the causal mechanisms between the built environment, occupants, and microbiomes, which will provide a knowledge base for the development of transformative intervention strategies toward healthy built environments. The pressing need to control the transmission of SARS-CoV-2 in indoor environments highlights the urgency and significance of understanding the complex interactions between the built environment, occupants, and microbiomes, which is the focus of this review.

Degradation of 17ß-estradiol by Novosphingobium sp. ES2-1 in aqueous solution contaminated with tetracyclines.

17ß-estradiol (E2) often coexists with tetracyclines (TCs) in wastewater lagoons at intensive breeding farms, threatening the quality of surrounding water bodies. Microbial degradation is vital in E2 removal, but it is unclear how TCs affect E2 biodegradation. This primary study investigated the mechanisms of E2 degradation by Novosphingobium sp. ES2-1 in the presence of TCs and assessed the removal efficiency of E2 by strain ES2-1 in natural waters containing TCs. E2 biodegradation was unaffected at TCs concentrations below 0.1 mg L-1 yet significantly inhibited at TCs above 10 mg L-1. As elevation of TCs, E2 biodegradation rate constant decreased, and the biodegradation kinetics equation gradually deviated from the pseudo-first-order dynamics model. Importantly, the presence of TCs, especially at high-level concentrations, significantly hindered E2 ring-opening process but promoted the condensation of some phenolic ring-opening products with NH3, thereby increasing the abundance of pyridine derivatives, which were difficult to decompose over time. Additionally, strain ES2-1 could remove 52.1-100% of nature estrogens in TCs-contaminated natural waters within 7 d. Results revealed the mechanisms of TCs in E2 biodegradation and the performance of a functional strain in estrogen removal in realistic TCs-contaminated aqueous solution.

Controlling the selective synthesis of [2]- and [3]rotaxanes by intermolecular steric hindrance between the macrocyclic hosts.

Two hydrogen-bonded azo-macrocycles with little disparity of the side chains in steric hindrance exhibited a substantial difference in complexation (slow/fast exchange) towards bipyridinium. Inspired by this finding, these macrocycles were applied to efficiently and selectively construct [2]- and [3]rotaxanes through one-pot synthesis. The origin of the selectivity in this novel approach was elucidated by comparing single crystal structures, DFT calculations and stepwise synthesis.

Theoretical Exploration of Carrier Dynamics in Amorphous Pyrene-Fluorene Derivative Organic Semiconductors.

In this report, a series of amorphous organic optoelectronic pyrene-fluorene derivative materials (BP1, BP2, PFP1, PFP2, OP1, OP2) were systematically investigated through a theoretical method. Their molecular structures are different due to the difference of substitution groups at C9 of the fluorene core, which include electron-rich pyrene group (PFP1 and PFP2), relatively neutral phenyl group (BP1 and BP2), and electron-withdrawing oxadiazole group (OP1 and OP2). In the beginning, through the physical model analysis, this report proposes that the concept of p-type or n-type is not flawless because there is no real doping process in these molecular organic semiconductors. To prove such a concept, the Marcus theory and first-principles were employed to calculate the intrinsic transfer mobility of these materials. Not as the common method used for the single crystal, in this report, a series of disorderly designed lattice cells were constructed to represent the disordered distribution of the amorphous pyrenyl-fluorene derivatives. Then, the reorganization energy of materials was calculated by the adiabatic potential energy surface method. The transfer integral of dimers was calculated in possible hopping pathways near the central molecule. Research results show that the six pyrene-fluorene materials all possess intrinsic bipolar transfer characteristics. In addition, it is also showed that the electron-rich group is not necessary to improve hole transfer, and that the electron-withdrawing group is also not necessary to improve electron transfer.

A Dynamic Hydrogen-Bonded Azo-Macrocycle for Precisely Photo-Controlled Molecular Encapsulation and Release.

A light-responsive system constructed from hydrogen-bonded azo-macrocycles demonstrates precisely controlled propensity in molecular encapsulation and release process. A significant decrease in the size of the cavity is observed in the course of the E→Z photoisomerization based on the results from DFT calculations and traveling wave ion mobility mass spectrometry. These macrocyclic hosts exhibit a rare 2:1 host-guest stoichiometry and guest-dependent slow or fast exchange on the NMR timescale. With the slow host-guest exchange and switchable shape change of the cavity, quantitative release and capture of bipyridinium guests is achieved with the maximum release of 68 %. This work underscores the importance of slow host-guest exchange on realizing accurate release of organic cations in a stepwise manner under light irradiation. The light-responsive system established here could advance further design of novel photoresponsive molecular switches and mechanically interlocked molecules.