SG-RAPL:
Scene Graph-Driven Reasoning
for Action Planning of Humanoid Robot

Yudin Dmitry · Lazarev Aleksandr . Bakaeva Eva
Kochetkova Angelika · Panov Aleksandr · Kovalev Aleksei

Project | Code

Description

This project is an open-source code base for our SG-RAPL project. The paper is submited to Engineering Applications of Artificial Intelligence.

Dataset

SG-RAPL Datset can be found here.

Launch

This project follows a microservice architecture, microservices communicate with each other with volumes.

To get started clone this repository:

git clone https://github.com/SashaDance/SG-RAPL.git

Run a service

docker compose up -d --build
export COMPOSE_FILE=compose.base.yaml:compose.dev.yaml
docker compose up -d --build $SERVICE

Test service

export COMPOSE_FILE=compose.base.yaml:compose.dev.yaml
docker compose up -d --build $SERVICE
docker compose exec $SERVICE bash -c "poetry run pytest -vv"