Event-based dataset for the detection and classification of manufacturing assembly tasks

TL;DR

This paper introduces EDAT24, a novel event-based dataset capturing manufacturing assembly tasks using a DAVIS240C sensor, aimed at advancing real-time detection and classification of human actions in industrial settings.

Contribution

The paper presents EDAT24, the first dataset of its kind for manufacturing primitives using event-based vision, including raw and pre-processed data, with tools for extension.

Findings

Dataset includes 400 samples of manufacturing primitives.

Data captured with DAVIS240C event camera for real-time analysis.

Tools provided for dataset extension and new primitive addition.

Abstract

The featured dataset, the Event-based Dataset of Assembly Tasks (EDAT24), showcases a selection of manufacturing primitive tasks (idle, pick, place, and screw), which are basic actions performed by human operators in any manufacturing assembly. The data were captured using a DAVIS240C event camera, an asynchronous vision sensor that registers events when changes in light intensity value occur. Events are a lightweight data format for conveying visual information and are well-suited for real-time detection and analysis of human motion. Each manufacturing primitive has 100 recorded samples of DAVIS240C data, including events and greyscale frames, for a total of 400 samples. In the dataset, the user interacts with objects from the open-source CT-Benchmark in front of the static DAVIS event camera. All data are made available in raw form (.aedat) and in pre-processed form (.npy). Custom-built Python code is made available together with the dataset to aid researchers to add new manufacturing primitives or extend the dataset with more samples.