Marine debris usually contains elements that are difficult to degrade and have a negative impact on the environment. Remote controlled underwater vehicles can remove trash from the deep ocean, but the efficiency of the detection algorithm used by these vehicles plays a significant role in its success. The proposed work describes the YOLOIncep Network, a novel and efficient deep-sea debris detection model. The proposed work uses the advantage of both YOLO and the inception Network. As a preprocessing step, a super resolution technique is used to improve the resolution of the image, thereby improving the detection performance of the debris. The effectiveness of the proposed work is evaluated using three categories from the publicly available JAMSTEC dataset. The results of our work are compared to other existing models, and it is observed that YOLOIncep outperforms other YOLO models, with a MAP @ 0.5 of 0.979.