The Map901 project is a collaborative effort between the City of Memphis and the University of Memphis to create accurate, accessible indoor maps and models that will improve the safety of the public and first responders.
Although the University of Memphis has already gathered data for these models by surveying 1.86 million square feet of indoor space, more data is needed to refine the indoor mapping model. The Map901 team is offering an exciting opportunity for students to contribute to public safety and computer vision research, while also winning prizes!
The Map901 team is hosting an image annotation contest that will give participants the opportunity to learn how to use an annotation tool to build image databases for computer vision research.
Image annotation is a technique of labeling images with certain outlines and keywords, or labels, in order to make them recognizable for machines. Image annotation can be done manually by humans using annotation tools or software. The keywords, or labels, are predetermined by the project engineers and are chosen to give the computer vision model information about what is shown in the image. Depending on the project and its goals, the type and number of labels in each image can vary. In the Map901 project, our team is focused on identifying and labeling public safety objects such as fire extinguishers, fire alarms, and building entrance and exits.
One of the main purposes of image annotation is to use these manually annotated images as training data sets in machine learning. It is important to annotate and label the images with the accurate object outlines and correct labels, so they are easily recognizable for machines.
- Contest Details
- Instructions for Image Annotation
If you are interested in learning more about the contest details or image annotation instructions, more information can be found at the links below.
