intelligent-control-lab / mekf_mame

Modified Extended Kalman Filter with generalized exponential Moving Average and dynamic Multi-Epoch update strategy (MEKF_MAME)

MEKF_EMA-DME

Modified Extended Kalman Filter with generalized Exponential Moving Average and Dynamic Multi-Epoch update strategy (MEKF_EMA-DME)

Pytorch implementation source coder for paper Robust Online Model Adaptation by Extended Kalman Filter with Exponential Moving Average and Dynamic Multi-Epoch Strategy.

In this paper, inspired by Extended Kalman Filter (EKF), a base adaptation algorithm Modified EKF with forgetting factor (MEKF_λ) is introduced first. Then using exponential moving average (EMA) methods, this paper proposes EMA filtering to the base EKF_λ in order to increase the convergence rate. In order to effectively utilize the samples in online adaptation, this paper proposes a dynamic multi-epoch update strategy to discriminate the “hard” samples from “easy” samples, and sets different weights for them. With all these extensions, this paper proposes a robust online adaptation algorithm: MEKF with Exponential Moving Average and Dynamic Multi-Epoch update strategy (MEKF_EMA-DME).

Requirements

• Python 3.6
• pytorch >=1.1.0
• pip install -r requirements.txt

How to use it

1 . Offline Neural Network Training

python train.py

2 . Online Adaptation

python adapt.py

You can online adapt the offline trained model with several optimizers, including SGD, Adam, MEKF_λ, MEKF_EMA-DME.