http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Evaluation of Recurrent Neural Network Variants for Person Re-identification
Cuong Vo Le,Nghia Nguyen Tuan,Quan Nguyen Hong,Hyuk-Jae Lee 대한전자공학회 2017 IEIE Transactions on Smart Processing & Computing Vol.6 No.3
Instead of using only spatial features from a single frame for person re-identification, a combination of spatial and temporal factors boosts the performance of the system. A recurrent neural network (RNN) shows its effectiveness in generating highly discriminative sequence-level human representations. In this work, we implement RNN, three Long Short Term Memory (LSTM) network variants, and Gated Recurrent Unit (GRU) on Caffe deep learning framework, and we then conduct experiments to compare performance in terms of size and accuracy for person reidentification. We propose using GRU for the optimized choice as the experimental results show that the GRU achieves the highest accuracy despite having fewer parameters than the others.
An Adaptive Supper Twisting Algorithm-based Terminal Sliding Mode Control for Robotic Manipulators
Van-Cuong Nguyen,Hee-Jun Kang,Anh-Tuan Vo,Thanh-Nguyen Truong 제어로봇시스템학회 2021 제어로봇시스템학회 국내학술대회 논문집 Vol.2021 No.6
In this paper, a continuous control method is proposed to deal with the effects of lumped uncertainties and faults of robot manipulators. A terminal sliding mode control technique is employed to track the expected trajectories. In order to decrease the chattering of sliding mode control, an adaptive super-twisting algorithm is performed. The proposed controller provides continuous control signal, finite-time convergence of tracking error, and robustness against the lumped uncertainties and faults. In addition, it improves the tracking error precision and reduces the chattering phenomenon. Moreover, the requirement of the lumped uncertainties and faults prior knowledge is removed using the adaptive law. Computer simulations are performed for a 2-link serial robot manipulator to confirm the effectiveness of the proposed algorithm.