ALL-IDB is a dataset of microscopic images of blood samples, specifically designed for the evaluation and the comparison of algorithms for segmentation ...
MPII Human Pose dataset is a state of the art benchmark for evaluation of articulated human pose estimation. The dataset includes around25K imagescontai...
The aim of the benchmark is to assess the accuracy and reliability of current methods for calibration and orientation of images acquired by different pl...
The Daimler Urban Segmentation Dataset consists of video sequences recorded in urban traffic. The dataset consists of 5000 rectified stereo image pairs ...
This work aims at analyzing the statistical properties of computer generated images by application of the Natural Scene Statistics (NSS) models. For the...
To advance the state of the art in leaf segmentation and to demonstrate the difficulty of segmenting all leaves in an image of plants, a challenge calle...
The de-facto image dataset for new algorithms. Many image API companies have labels from their REST interfaces that are suspiciously close to the 1000 c...
natural-imageScene understanding with many ancillary tasks (room layout estimation, saliency prediction, etc.) and an associated competition.
natural-image