Working memory (WM) is a fundamental cognitive ability to support complex thought, but it is limited in capacity. WM training has shown the potential to make benefit for those in need. Many studies have shown the potential of transcranial direct current stimulation (tDCS) to transiently enhance WM performance by delivering low current to the brain cortex of interest via electrodes on the scalp. TDCS has also been revealed as a promising intervention to augment WM training in a few studies. However, (...) those few tDCS-paired WM training studies focused more on the effect of tDCS on WM enhancement and its transferability after training but paid less attention to the variation of cognitive performance during the training procedure. The current study attempted to explore the effect of tDCS on the variation of performance during WM training in healthy young adults. All the participant received WM training with the load-adaptive verbal N-back task for five days. During the training procedure, active/sham anodal high-definition tDCS (HD-tDCS) was used to stimulate the left dorsolateral prefrontal cortex (DLPFC). To examine the training effect, pre- and post-test were performed respectively one-day before and after the training sessions respectively. At the beginning of each training session, stable-load WM tasks were performed to examine the performance variation during training. Compared to the sham stimulation, higher learning rates of performance metrics during the training procedure were found when WM training was combined with active anodal HD-tDCS. The performance improvements (post-pre) of the active group were also found to be higher than those of sham group and were transferred to a similar untrained WM task. Further analysis revealed that the negative relation between the training improvements and baseline performance. These findings show the potential that tDCS may be leveraged as an intervention to facilitate WM training for those in need of higher WM ability. (shrink)

Real-time smoke detection is of great significance for early warning of fire, which can avoid the serious loss caused by fire. Detecting smoke in actual scenes is still a challenging task due to large variance of smoke color, texture, and shapes. Moreover, the smoke detection in the actual scene is faced with the difficulties in data collection and insufficient smoke datasets, and the smoke morphology is susceptible to environmental influences. To improve the performance of smoke detection and solve the problem (...) of too few datasets in real scenes, this paper proposes a model that combines a deep convolutional generative adversarial network and a convolutional neural network to extract smoke features and detection. The vibe algorithm was used to collect smoke and nonsmoke images in the dynamic scene and deep convolutional generative adversarial network used these images to generate images that are as realistic as possible. Besides, we designed an improved convolutional neural network model for extracting smoke features and smoke detection. The experimental results show that the method has a good detection performance on the smoke generated in the actual scenes and effectively reduces the false alarm rate. (shrink)