Skip to content

Latest commit

 

History

History
42 lines (33 loc) · 1.21 KB

README.md

File metadata and controls

42 lines (33 loc) · 1.21 KB

ANAVI

ANAVI is a framework for "Audio Noise Awareness using Visuals of Indoor environments for NAVIgation". This is presented at 8th Annual Conference on Robot Learning, 2024.

Website: https://anavi-corl24.github.io

Setup

  1. Download or clone this repo
  2. Creat conda/mamba environment
  3. Install anp by: pip install -e .
  4. Install PyTorch

Generate Data

Follow instructions on DATAGEN.md. Alternately, you can also download the simulated data sample here.

Run

  1. Train
python train.py --config=dirdis_huberloss.yaml
python train.py --config=ego_huberloss.yaml
python train.py --config=pano_huberloss.yaml

You can train your own model by modifying config_template.yaml.

  1. Evaluate python train.py --config=configs/visdirdis_ce.yaml

Cite

@article{jain2024anavi,
  author    = {Jain, Vidhi and Veerapaneni, Rishi and Bisk, Yonatan},
  title     = {ANAVI: Audio Noise Awareness using Visuals of Indoor environments for NAVIgation},
  journal   = {8th Annual Conference on Robot Learning},
  year      = {2024},
}

Contact Us

If you have questions or concerns, you can raise an issue or email vidhij@andrew.cmu.edu.