Finding images for your machine learning project can take a long time. Especially if you want to create a diverse dataset that also uses images with a correct license. On CrowdCapture you can create your own projects or contribute to other open source datasets. Every image uploaded is checked by other users. Every week a new version of the dataset will be available to download. An example of an existing dataset can be found here: https://crowdcapture.org/project/000fb108-97a0-44b0-b3f8-ba3139848a84