image
imagewidth (px)
500
500
label
class label
2 classes
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image
0image

image/jpeg

The dataset contains 500 by 500 pixel images with a green border around them. Each image is accompanied by a mask represented as a black and white image. In this mask, white color highlights the regions where the person in the image is present and black color denotes the rest of the regions. The dataset provides an opportunity to investigate the detection and segmentation of people in the images. The images have been selected and processed to have a uniform size and to be framed in green color. This provides a convenient basis for developing and testing object detection algorithms. Each image has a corresponding mask associated with it, which helps in identifying the pixels belonging to the person in the photograph. This is useful for object segmentation tasks such as selecting regions containing a person for further image analysis and processing.

Downloads last month
48
Edit dataset card