Snapshot Serengeti
A large-scale camera-trap wildlife image dataset containing annotated images of multiple animal species in natural scenes, commonly used in wildlife recognition and image reconstruction research.
Timeline (last 90 days)
On 4,744 test images from the Snapshot Serengeti dataset, reconstructed at a 75% masking ratio, WildMAE achieves a PSNR of 27.90 dB, compared with 24.33 dB for the original MAE.
Snapshot Serengeti is one of the most representative public datasets in the camera-trap field, consisting of hundreds of thousands of images taken over many years by hundreds of cameras in Serengeti National Park.
All Facts (2)
On 4,744 test images from the Snapshot Serengeti dataset, reconstructed at a 75% masking ratio, WildMAE achieves a PSNR of 27.90 dB, compared with 24.33 dB for the original MAE.
50%UnverifiedSnapshot Serengeti is one of the most representative public datasets in the camera-trap field, consisting of hundreds of thousands of images taken over many years by hundreds of cameras in Serengeti National Park.
50%