Deep convolutional feature aggregation for fine-grained cultivar recognition

Hao Wu, Lincong Fang, Qian Yu, Chengzhuan Yang

Abstract: Highlights•We propose a regional convolutional covariance feature (RCCF) description method.•The RCCF method improves the identification ability of fine-grain cultivars.•We present a multiresolution regional maximum activation of convolutions method.•Our method achieves state-of-the-art performance on benchmark cultivar dataset.