Learned image compression is making good progress in recent years. Peak signal-to-noise ratio (PSNR) and multiscale structural similarity (MS-SSIM) are the two most popular evaluation metrics. As different metrics only reflect certain aspects of human perception, works in this field normally optimize two models using PSNR and MS-SSIM as loss function separately, which is suboptimal and makes it difficult to select the model with best visual quality or overall performance. Towards solving this problem, we propose to Harmonize optimization metrics in Learned Image Compression (HLIC) using online loss function adaptation by reinforcement learning. By doing so, we are able to leverage the advantages of both PSNR and MS-SSIM, achieving better visual quality and higher VMAF score. To our knowledge, our work is the first to explore automatic loss function adaptation for harmonizing optimization metrics in low level vision tasks like learned image compression.

Multi-car elevator system is a new technique which can reduce the occupied area and improve the transport efficiency in the high-rise and super high-rise buildings. It is remain to be studied in both the control strategy and scheduling method compared with the traditional single elevator group control system, In order to improve the utilization rate and shorten the waiting time, the particle swarm algorithm PSO and GA genetic algorithm optimization control method are studied aiming at the rationality of scheduling strategy of the rationality of elevator. Combining the advantages of two algorithms proposed a novel optimal scheduling algorithm for PSO-GA and carried on simulation analysis, the experimental results demonstrate the effectiveness of the new method of improved PSO-GA.

This paper focuses on the issues tower crane safety hard to be comprehensive evaluated, the machine learning theory is introduced to the tower crane safety analysis. Firstly, collect the tower crane operating parameters as samples, the SVM classifier make supervised-learning on the samples data, established classification model and running parameters are used to evaluate the secure state. In addition, the effect of different kernel functions and parameters on SVM classifier is discussed, parameters are selected after optimization by mixed method of cross-validation and grid search. Experiments show that optimized SVM model can judge Tower crane safety correctly.

No abstract

This paper presents a method based on adaptive two-dimensional wavelet packet for three-phase power quality data. In this paper, we shall briefly introduce the method of dq0 transformation. First, the original data is converted by the method of dq0 transformation, its purpose may be to eliminate redundancy between the three-phases alternating data, then change the one-dimensional date converted into a two-dimensional matrix by the integral multiple of cycles, which aims to eliminate redundant between cycles. Finally, use the method of two-dimensional adaptive wavelet packet decomposition, image SPIHT coding and DEFLATE coding to compress the data. The simulation result shows that the algorithm effectively compress the three-phase power quality data, and it has a good compression effect.