A TextGCN-Based Decoding Approach for Improving Remote Sensing Image Captioning
Swadhin Sambit Das, Raksha Sharma · IEEE Geoscience and Remote Sensing Letters · 2024
Remote sensing (RS) images are highly valued for their ability to address complex real-world issues such as risk management, security, and meteorology. However, manually captioning these images is challenging and requires specialized knowledge across various domains. This letter presents an approach for automatically describing (captioning) RS images. We propose a novel encoder-decoder (ED) setup that deploys a text graph convolutional network (TextGCN) and multilayer long short-term memory (LSTM). The embeddings generated by TextGCN enhance the decoder’s understanding by capturing the semantic relationships among words at both the sentence and corpus levels. Furthermore, we advance our approach with a comparison-based beam search method to ensure fairness in the search strategy for generating the final caption. We present an extensive evaluation of our approach against various other state-of-the-art ED frameworks. We evaluated our method using seven metrics: BLEU-1 to BLEU-4, METEOR, ROUGE-L, and CIDEr. The results demonstrate that our approach significantly outperforms other state-of-the-art ED methods.