The main focus mechanism is an efficient vehicle for learning another feature recognition tasks embedding method, the region of interest can be obtained through the training we need depth learning network. <br><br>Can be found, most of the attention-based model will focus concentrated on the part of the region, when a vehicle when the image background is too complex will cause poor results.
正在翻译中..