Wearable spoken guidance with task-aware visual content selection and inertial gating. NavGuide includes a local device service, tactile paving and street crossing workflows, item search, an optional cloud description gateway, and XIAO ESP32S3 Sense firmware.
Use Python 3.11:
git clone https://github.com/InSAI-Lab/NavGuide.git
cd NavGuide
python3.11 -m venv .venv
source .venv/bin/activate
python -m pip install -r requirements-dev.txt -c constraints.txt
cp .env.example .env
python -m navguideOpen http://127.0.0.1:8081. The default observations mode accepts detection results without loading a vision model. Camera navigation requires the models and device setup described below.
Run tests with python -m pytest -q.
- Local deployment, device interfaces, and field validation
- Cloud gateway deployment
- Hardware wiring, firmware configuration, and flashing
- Model files and dependencies
- Algorithm, parameter provenance, and evaluation
- Project structure
For extended navigation, install requirements-navigation.txt, prepare the task models and audio assets, then run python -m navguide.navigation.app.
See LICENSE for license terms.