A Novel Exposure Fusion Method Based on Low-Resolution Context Aggregation Attention Network

Recently, deep learning based multi-exposure image fusion methods have been widely explored due to their high efficiency and adaptability. In this paper, we propose a multi-exposure image fusion method based on low-resolution context aggregation attention network (CAAN). First, we feed the low-resolution version of the input images to CAAN to predict their low-resolution weight maps. Then, the high-resolution weight maps are generated by guided filtering for upsampling (GFU). Finally, the high-resolution fused image is generated by weighted summation operation. Our proposed network is unsupervised and adaptively adjust the weights of channels to achieve better feature extraction. Experimental results show that our method outperforms existing state-of-the-art methods by both quantitative and qualitative evaluation.