IEEE ICIP 2023 CHALLENGE ON THE AUTOMATIC DETECTION OF MOSQUITO BREEDING GROUNDS

Every year, millions of people are infected with diseases transmitted by the Aedes aegypti mosquito. Preventing vector reproduction by identifying and eliminating possible breeding grounds is still the primary method to combat these diseases. In this challenge, we present a comprehensive dataset of aerial videos containing potential mosquito breeding sites acquired using drones. All objects of interest were manually annotated with bounding boxes: discarded tires, water tanks, buckets, puddles, pools, and bottles. Proposed systems shall locate these items at both frame and individual-object levels. Participants from three different countries (namely Brazil, Qatar, and Spain) submitted results, which may be compared to previous ones found in the related literature.