Using convolutional neural network, the velocity of the camera is estimated. After predicting the velocity, 3D transformation matrices are concanated to estimate the position. Yet, the rotation information is coming from gyro scope. The correction is also done with accelerometer
Hongyun Lee, Matthew McCrink, and James W. Gregory. "Visual-Inertial Odometry for Unmanned Aerial Vehicle using Deep Leering", 2019 Intelligent/Autonomous Guidance and Navigation, AIAA SciTech Conference, accepted for publication
- DeepVO: Towards End-to-End Visual Odometry with Deep Recurrent Convolutional Neural Networks(https://senwang.gitlab.io/DeepVO/files/wang2017DeepVO.pdf)
- LSTM integration
- upload weight.pt
- explain data set & location
pip install numpy pip install scipy pip install pandas pip install matplotlib pip install scikit-learn pip install pathlib pip install pypng pip install pillow pip install django pip install image pip install opencv-python opencv-contrib-python
detail: https://github.com/ElliotHYLee/SystemReady
- Hardware
CPU: i9-7940x RAM: 128G, 3200Hz GPU: two Gefore 1080 ti MB: ROG STRIX x299-E Gaming
- Software
Windows10 Python3 with native pip PyTorch: v1 CUDA: v9.0 Cudnn: v7.1
python main_cnn.py
description
