Ovarian tumours are abnormal growths on the ovaries which can be either cancerous(malignant) or non-cancerous(benign).In case of malignancy, the cancer cells are most likely to grow quickly and get out of control within a year. This is why early diagnosis plays a key role in the survival of ovarian cancer. The traditional way of identifying malignant and benign tumours is biopsy. However, with the advancement of technology, computer-aided methods have emerged for such diagnostic purposes. This work is one such computer-aided method to effectively detect both benign and malignant tumours using Detectron2 Deep Learning network.The performance of the network is analysed using various parameters such as Dice-coefficient, precision, recall. The network has yielded an accuracy of 98.3% and the observations are elaborated in the results and discussion section. © 2023 Elsevier B.V., All rights reserved.