Official page of RONet, which is published @IROS'19
Since the original code is based on Tensorflow, now I have ported the original algorithm to PyTorch.
- Port RONet
- Port Bi-LSTM
- Run on test data
- Set training pipeline
- Visualize training procedure
- Autosave the best model
Please refer to requirements.txt
UWB is abbv. for Ultra-wideband, and the sensor outputs only 1D range data.
More explanations are provided in this paper.
In summary, UWB data are likely to be vulnerable to noise, multipath problems, and so forth.
Thus, we leverage the nonlinearity of deep learning to tackle that issue.
All data are contained in uwb_dataset and a total of eight sensors are deployed, whose positions are as follows:
And each csv consists N (the num. of sequences) x 10 whose columns denotes:
range @id0, range @id1, range @id2, range @id3, range @id4, range @id5, range @id6, range @id7, x of GT, y of GT
Note that our experiment was conducted on real-world data by using Pozyx UWB sensors and the motion capture system.
(Please kindly keep in mind that Pozyx systems do not give precise range data :( )
The training scripts come with several options, which can be listed with the --help flag.
python3 main.py --helpThe point is that it only takes a few minutes because the data of UWB are lightweight and simple! :)
Training results will be saved under the results folder. To resume a previous training, run
python3 main.py --resume [path_to_previous_model]python3 main.py --evaluate [path_to_trained_model]On validation data
| Methods | RMSE (cm) |
|---|---|
| RNN | 4.050 |
| GRU | 3.918 |
| LSTM | 4.855 (what's wrong with you..?) |
| Bi-LSTM | TBA |
| RONet | TBA |
- If you use our code or method in your work, please consider citing the following::
@INPROCEEDINGS {lim2019ronet,
author = {Lim, Hyungtae and Park, Changgue and Myung, Hyun},
title = {Ronet: Real-time range-only indoor localization via stacked bidirectional lstm with residual attention},
booktitle = {Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
pages={3241--3247},
year = { 2019 },
organization={IEEE}
}
@INPROCEEDINGS {lim2018stackbilstm,
author = {Lim, Hyungtae and Myung, Hyun},
title = {Effective Indoor Robot Localization by Stacked Bidirectional LSTM Using Beacon-Based Range Measurements},
booktitle = {International Conference on Robot Intelligence Technology and Applications},
pages={144--151},
year = { 2018 }
organization={Springer}
}
Contact: Hyungtae Lim ([email protected])
Please create a new issue for code-related questions. Pull requests are welcome.