DeepFake detection has become an attractive research topic with tremendous growth of interests recently. However, existing DeepFake detection studies spare no effort to improve accuracy or Area Under Curve metric, regardless of computing costs. In this work, the tradeoff between result accuracy and computing resources is taken into consideration. A facial sparse optical flow method is proposed to extract spatio-temporal features representing facial expression incoherence, which helps to distinguish fake videos and real videos. The features fed into a light CNN model are highly compact and low-dimensional. The proposed method has an amazing small amount of parameters with high training speed and low usage of GPU memory. The low resource requirement makes it possible to port to embedded development platform.

The bearing preload has significant influence on dynamic performance of motorized spindle. The dynamic performance and the life of bearing can be significant improve if the bearing preload can change according to the different conditions. The main purpose of this paper is to calculate the minimum bearing preload to prevent gyroscopic rotation and study the natural frequency and mode analysis of spindle vibration. Perform the experiments of variable bearing preload device based on piezoelectric actuators. Research the mechanical characteristic of piezoelectric actuators. Providing theory for improving the performance of the motorized spindle by controlling the bearing preload .

No abstract

No abstract