SAWIT

SAWIT (small-sized animal wild image dataset) is the first real-world dataset of small-sized animals, collected from camera traps and in realistic conditions. This dataset consists of 34,434 images and is annotated by experts in the field with object-level annotations (bounding boxes) providing 34,820 annotated animals for seven animal categories: frog, lizard, bird, small mammal, medium or big mammal, spider, and scorpion. The dataset encompasses a wide range of challenging scenarios, such as occlusions, blurriness, and instances where animals blend into the dense vegetation. Based on the dataset, we benchmark two prevailing object detection algorithms: Faster RCNN and YOLO, and their variants. Our paper can be found at: https://link.springer.com/article/10.1007/s11042-023-16673-3.

Citation

If you use our provided data and/or models, please cite our work at,

@article{SAWIT,
    author       = {Thi Thu Thuy Nguyen and 
                  Anne C. Eichholtzer and 
                  Don A. Driscoll and 
                  Nathan I. Semianiw and 
                  Dean M. Corva and 
                  Abbas Z. Kouzani and 
                  Thanh Thi Nguyen and 
                  Duc Thanh Nguyen},      
    title        = {SAWIT: A small-sized animal wild image dataset with annotations}, 
    journal      = {Multimedia Tools and Applications},
    year         = {2023} 
}