Video Abnormal Event Detection Based on One-Class Neural Network.

Video abnormal event detection is a challenging problem in pattern recognition field. Existing methods usually design the two steps of video feature extraction and anomaly detection model establishment independently, which leads to the failure to achieve the optimal result. As a remedy, a method based on one-class neural network (ONN) is designed for video anomaly detection.

In situ H(+)-mediated formation of singlet oxygen from NaBiO3 for oxidative degradation of bisphenol A without light irradiation: Efficiency, kinetics, and mechanism.

Bisphenol A (BPA) is a ubiquitous environmental contaminant with endocrine disruption potential. This study explored the efficiency, kinetics, and mechanism of BPA removal from weakly acidic solutions by using NaBiO3 as a source of singlet oxygen. It was observed that the use of NaBiO3 (1gL(-1)) could eliminate almost all (more than 97%) of the added BPA (0.