 ASAR dataset has been created consisting of 43,819 shipchips of 256 pixels in both range and azimuth. The dataset was labeled by SAR experts and contains ships with distinct scales and backgrounds. Additionally, modified state-of-the-art object detectors from natural images were trained and tested on the dataset. The results showed that these object detectors achieved higher mean average precision, map, on the test dataset and had high generalization performance on new SAR imagery without land-ocean segmentation. This article was authored by Yuan Yuan Wang, Chao Wang, Hong Zeng, and others.