MosquitoFusion Dataset

Preprocessing

• Standardization: Images are auto-oriented and resized to 640×640 pixels.
• Annotation Integrity: All images contain relevant annotations, adhering to a strict filter null criterion.

Augmentation

• Techniques Used: Flips, rotations, crops, and grayscale applications.
• Purpose: To enhance dataset diversity, promoting robust model training.

The dataset includes a wide variety of scenarios, capturing mosquitoes in different environments and conditions. This ensures the models trained on it are well-rounded and effective in diverse real-world applications. By providing high-quality, annotated images, MosquitoFusion stands as a vital resource for researchers and developers aiming to improve mosquito detection technologies, which are crucial for public health and disease control.