Deep learning (DL) has changed many areas, especially in computer vision and natural language processing. Traditional DL algorithms work well with data that has a Euclidean structure, like images, text, and audio, using several non-linear layers to find important features. However, they struggle with non-Euclidean data, especially data that is structured as graphs. This issue has led researchers to look for new ways to effectively integrate graph data into DL frameworks. The development of Graph Neural Networks (GNNs) is a key advancement in artificial intelligence (AI), aiming to connect the strong feature extraction abilities of deep learning with the complex relationships in graph data. The creation of GNNs has sparked significant interest in their use across many fields because they can model complex structures and relationships. By representing data as graphs, GNNs maintain the connections between data points, allowing for more detailed analyses and conclusions. In computer vision, this results in better object detection, action recognition, and scene understanding, as GNNs effectively manage the spatial and temporal connections that traditional DL models may miss. Additionally, GNNs are recognized as powerful tools in cross-media tasks like visual question answering and image captioning, where the complex interaction of contextual information is critical for improving model accuracy and understanding. Importantly, the rise of GNNs not only suggests improved performance but also offers chances for better robustness and explains ability in AI systems. By using graph representations of data, GNNs can clarify their decision-making processes, which is essential for applications where understanding model behaviour is important. As these networks continue to develop, they are poised to change many business needs, allowing for the creation of advanced applications that were not possible before. The path to fully harnessing the potential of GNNs in computer vision and other areas is just starting, paving the way for future research to explore their abilities, improve their designs, and use their strengths to tackle complex problems in various industries.