This is the project folder for the journal articel "Active head rolls enhance sonar-based auditory localization performance"
This project provide data and code related to the journal pulblication. The provided codes form the Active Efficient Coding (AEC) model.
Here, we will describe the main codes required for training and testing the proposed model.
| Code | Description |
|---|---|
| Main.m | Main file required for training |
| Generate_test_trials.m | Generate testing trials |
| Model.m | Model class |
| ASSOMOnline.m | GASSOM |
| CActorG.m | Actor (Reinforcement Learning) |
| CCritic.m | Critic(Reinforcement Learning) |
HRTF data
We are grateful to the research group headed by Prof. Cynthia F. Moss at Department of Psychological & Brain Sciences, Johns Hopkins University, USA for the HRTF data access. We specially acknowledge Dr. Murat Aytekin for all the support with the data. The HRTF data is organized by the subject number indicated by the name HRTF_subject#.mat. Each file contain the direction of sound source and also the left and right HRTFs. Each HRTF is in the frequency domain and the coefficients are in complex form for each frequency bin. The raw data used to compute the HRTFs are not provided with this dataset. Please contact Prof. Cynthia F. Moss using the email cynthia.moss@gmail.com to request the raw data.
To run this project, download the project folder and run:
Main
Generate_test_trials
The program "Main.m" will save the learned model to the folder "Subject_{subject number}trail{trial number}sigma{roll angle standard deviation}". "Generate_test_trials.m" will generate the tresting trial trajectories and plot the trajectories.
All the codes are tested with the MATLAB version R2017a (64-bit) under ubuntu 16.04 LTS.
We would like to acknowledge the graduate students contributed to the Active Efficient Coding (AEC) model over the years under the supervision of Prof. Jochen Triesch and Prof. Bertram E. Shi.
- Yu Zhao
- Thusitha N. Chandrapala
- Chong Zhang
- Qingpeng Zhu
- Céline Teulière
- Luca Lonini
