A very large free repository of image data designed for use in computer vision and object recognition research consisting of over 14 million labeled images. Created in 2006 by American researcher Fei-Fei Li, ImageNet was made into a significant object recognition challenge for computer vision researchers. The ImageNet Large Scale Visual Recognition Challenge (ILSVRC), created in 2010, provided an opportunity for computer vision researchers to compete to correctly detect and classify objects and scenes. The ImageNet competition is undergoing changes to format and challenge subsequent to researchers achieving significantly accurate results, better than human performance, as well as concerns over the accuracy of ImageNet data with some sources stating that about 5% of the data is mis-labeled and contains embedded bias.
Menu
