We develop a novel nL-sized microdroplet laser based on the capillary optofluidic ring resonator (OFRR). The microdroplet is generated in a microfluidic channel using two immiscible fluids and is subsequently delivered to the capillary OFRR downstream. Despite the presence of the high refractive index (RI) carrier fluid, the lasing emission can still be achieved for the droplet formed by low RI solution. The lasing threshold of 1.54 µJ/mm(2) is achieved, >6 times lower than the state-of-the-art, thanks to the high Q-factor of the OFRR. Furthermore, the lasing emission can be conveniently coupled into an optical fiber. Finally, tuning of the lasing wavelength is achieved via highly efficient fluorescence resonance energy transfer processes by merging two different dye droplets in the microfluidic channel. Versatility combined with improved lasing characteristics makes our OFRR droplet laser an attractive platform for high performance optofluidic lasers and bio/chemical sensing with small sample volumes.

Using the convolutional neural network (CNN) method for image emotion recognition is a research hotspot of deep learning. Previous studies tend to use visual features obtained from a global perspective and ignore the role of local visual features in emotional arousal. Moreover, the CNN shallow feature maps contain image content information; such maps obtained from shallow layers directly to describe low-level visual features may lead to redundancy. In order to enhance image emotion recognition performance, an improved CNN is proposed in this work. Firstly, the saliency detection algorithm is used to locate the emotional region of the image, which is served as the supplementary information to conduct emotion recognition better. Secondly, the Gram matrix transform is performed on the CNN shallow feature maps to decrease the redundancy of image content information. Finally, a new loss function is designed by using hard labels and probability labels of image emotion category to reduce the influence of image emotion subjectivity. Extensive experiments have been conducted on benchmark datasets, including FI (Flickr and Instagram), IAPSsubset, ArtPhoto, and Abstract. The experimental results show that compared with the existing approaches, our method has a good application prospect.