top of page

ACID: Alberta Construction Image Dataset

The Alberta Construction Image Dataset (ACID) is a construction machine detection dataset developed by AIRCon-Lab at the University of Alberta. The images in ACID were collected from construction sites that are from all over the world. ACID focuses on integrating deep learning technology in the construction industry and can be downloaded from the University of Alberta library. It is the resource for the community to develop deep learning applications in the construction automation field. 

​

​

​

​

​

raw_dataset.png

There are ten types of machines in the ACID dataset, which are excavator, compactor, dozer, grader, dump truck, cement truck, wheel loader, backhoe loader, tower crane, and mobile crane. These machines are common in modern construction sites. ACID includes 10,000 labeled images, 15,767 construction machine objects for training deep learning detection algorithms.

​

More information about ACID can be acquired in the ACID website (www.acidb.ca)

​

bottom of page