This work proposes an unsupervised joint alignment framework, referred to as “Gradient Correlation Congealing,” which aligns an image ensemble by maximizing a sum of gradient correlation coefficient function defined over all images. We, respectively, develop two different formulations to optimize the objective function regarding the role of “template.” While most existing face alignment methods suffer from outliers, e.g., occlusions, the proposed algorithms are able to align faces undergoing partial occlusions. Moreover, our algorithms can cope with nonuniform illumination changes (even extremely difficult ones), and also, they do not require any predefined templates. We test the novel approaches against four typical joint alignment methods including Least-Squares Congealing, Learned-Miller Congealing, Lucas–Kanade entropy Congealing, and RASL using three challenging face databases: AR, Yale B, and LFW. Experimental results prove the efficiency of our approaches under different conditions, especially when faces are partially occluded, and the proposed algorithms perform much better than all considered methods. 相似文献
Journal of Computer Virology and Hacking Techniques - With macOS increasing popularity, the number, and variety of macOS malware are rising as well. Yet, very few tools exist for dynamic analysis... 相似文献
Edge caching has received much attention as a promising technique to overcome the stringent latency and data-hungry challenges in the future generation wireless networks. Meanwhile, full-duplex (FD) transmission can potentially double the spectral efficiency by allowing a node to receive and transmit at the same frequency band simultaneously. In this paper, we investigate the delivery time performance of a cache-aided FD system, in which an edge node, operates in FD mode, serves users via wireless channels and is equipped with a cache memory. Firstly, we derive a closed-form expression for the average delivery time by taking into account the uncertainties of both backhaul and access wireless channels. The derived analysis allows the examination of the impact of key parameters, e.g., cache size and transmit power. Secondly, a power optimization problem is formulated to minimize the average delivery time. To deal with the non-convexity of the formulated problem, we propose an iterative optimization algorithm based on the bisection method. Finally, numerical results are presented to demonstrate the effectiveness of the proposed algorithm. A significant delivery time reduction is achieved by the proposed optimization compared to the FD reference and half-duplex counterpart.
This paper represents the elements and the use of the upgraded simulation system, developed in the last half decade for ?tore Steel billet caster. The simulation system is used in the context of the state-of-the-art automation and information of the twenty-five year-old three-strand Concast billet continuous caster for dimensions square 140 and 180 mm with the capacity of 160,000 tons/year. The simulation system is used in the off-line and on-line modes. The off-line mode is used in order to set the proper process parameters and to calculate the temperature field, macrosegregation, and grain structure of the strand. It is also used to calculate the changes in the caster design such as the secondary cooling and the position of the SEN. The on-line model is used in automatic casting control system. The paper represents an update of our BHM publication of 2005 (Application of Continous Casting Simulation at ?tore Steel, BHM, Vol. 150, No. 9, 300–306). 相似文献
Multimedia Tools and Applications - Violent scenes detection (VSD) is a challenging problem because of the heterogeneous content, large variations in video quality, and complex semantic meanings of... 相似文献