Computer vision and deep learning are intrinsic applications in the machine learning field that become smarter day by day. The significant challenge in deep learning tasks converges on extracting the deepest and the most semantic features of the image. So, the promotion of deep learning techniques has enormous leverage in the retrieval of deep image ranking. This research has tackled an essential task of relevant deep image ranking based on learning from the RetinaNet superior detection technique. RetinaNet learning technique is employed to learn deep semantic features embeddings from the imaging dataset. Transfer of learning is a powerful scheme that proposes hyper-parameterization of the RetinaNet network for relevant image ranking. It transfers RetinaNet detector learning (weights) for deep relevant image ranking systems. Thus, we achieved the best accuracy. Our experimental results manifest that our deep learning procedure enhances the retrieval results efficiently and accurately and focuses on inhibiting the learning time of a deep relevant ranking task. As compared with other state-of-the-art object detectors, the RetinaNet detector accomplished more than a 97% mean average precision (MAP). This results in outperformed tested work. These superior results pretend the effective impact of our proposed procedure learning that drives the more efficient and relevant result of the deep ranking task.