Real-Time Illumination Invariant Face Detection Using Biologically Inspired Feature Set and BP Neural Network

Aiming at low precision for traditional angular velocity algorithms in GFSINS, a BP neural network algorithm without complex mathematic computation is put forward to calculate angular velocity. Based on a ten-accelerometer configuration scheme, the accelerometer output, sample interval and fixed position are chosen as input, angular velocity got by lognormal algorithm is chosen as output, and 5000 sample data is trained in the several conditions with different hiding layers, neural cells and training steps. Then a three-layer BP network model with 30 hiding layer neural cells is built. Finally, the angular velocity is predicted in real time by the model. Results show that network has strong adaptive capability and real time, and compared with lognormal algorithm, prediction time is almost equal, but prediction precision of angular velocity is nearly improved by three times.

Download Full-text

Bp Neural Network Implementation On Real-time Reconfigurable FPGA System For A Soft-sensing Process

2005 International Conference on Neural Networks and Brain ◽

10.1109/icnnb.2005.1614779 ◽

2006 ◽

Cited By ~ 6

Author(s):

Zhuo Ruan ◽

Jianguo Han ◽

Yuzhang Han

Keyword(s):

Neural Network ◽

Real Time ◽

Bp Neural Network ◽

Soft Sensing

Download Full-text

Real-Time Brain-Machine Interface Achieves High-Velocity Prosthetic Finger Movements using a Biologically-Inspired Neural Network Decoder

10.1101/2021.08.29.456981 ◽

2021 ◽

Author(s):

Matthew S. Willsey ◽

Samuel R. Nason ◽

Scott R. Ensel ◽

Hisham Temmar ◽

Matthew J. Mender ◽

...

Keyword(s):

Neural Network ◽

Real Time ◽

Finger Movements ◽

Feed Forward Neural Network ◽

Current Standard ◽

Biologically Inspired ◽

Current State ◽

The Neural Network ◽

Starting Point ◽

Prosthetic Finger

AbstractDespite the rapid progress and interest in brain-machine interfaces that restore motor function, the performance of prosthetic fingers and limbs has yet to mimic native function. The algorithm that converts brain signals to a control signal for the prosthetic device is one of the limitations in achieving rapid and realistic finger movements. To achieve more realistic finger movements, we developed a shallow feed-forward neural network, loosely inspired by the biological neural pathway, to decode real-time two-degree-of-freedom finger movements. Using a two-step training method, a recalibrated feedback intention–trained (ReFIT) neural network achieved a higher throughput with higher finger velocities and more natural appearing finger movements than the ReFIT Kalman filter, which represents the current standard. The neural network decoders introduced herein are the first to demonstrate real-time decoding of continuous movements at a level superior to the current state-of-the-art and could provide a starting point to using neural networks for the development of more naturalistic brain-controlled prostheses.

Download Full-text