Crowdsourcing Annotations for Visual Object Detection