Semantic Segmentation with RNNs and LSTMs: A Review of ReSeg and Brain Image Segmentation Techniques

In 2016, Visin et al. [25] proposed the ReSeg network, which introduced the ReNet model composed of RNNs, suitable for various semantic segmentation tasks. In 2018, Gao et al. [26] combined LSTM and CNN to model the temporal relationship between different brain MR slice images, achieving segmentation of brain images.