File(s) under permanent embargo

Reason: The file associated with this record is under embargo until publication, in accordance with the publisher's self-archiving policy. The full text may be available through the publisher links provided above.

Underwater object detection using Invert Multi-Class Adaboost with deep learning

conference contribution
posted on 04.05.2020, 15:14 by L Chen, Z Liu, L Tong, Z Jiang, S Wang, J Dong, Huiyu Zhou

Abstract—In recent years, deep learning based methods have achieved promising performance in standard object detection. However, these methods lack sufficient capabilities to handle underwater object detection due to these challenges: (1) Objects in real applications are usually small and their images are blurry, and (2) images in the underwater datasets and real applications accompany heterogeneous noise. To address these two problems, we first propose a novel neural network architecture, namely Sample-WeIghted hyPEr Network (SWIPENet), for small object detection. SWIPENet consists of high resolution and semantic-rich Hyper Feature Maps which can significantly improve small object detection accuracy. In addition, we propose a novel sample-weighted loss function which can model sample weights for SWIPENet, which uses a novel sample re-weighting algorithm, namely Invert Multi-Class Adaboost (IMA), to reduce the in-fluence of noise on the proposed SWIPENet. Experiments on two underwater robot picking contest datasets URPC2017 andURPC2018 show that the proposed SWIPENet+IMA framework achieves better performance in detection accuracy against several state-of-the-art object detection approaches.


Thanks for National Natural Science Foundation of Chinaand Dalian Municipal People’s Government providing theunderwater object detection datasets for research purposes.This project of underwater object detection is supported byChina Scholarship Council.



Proceedings of International Joint Conference on Neural Networks (2020) In Press


International Joint Conference on Neural Networks

Published in

Proceedings of International Joint Conference on Neural Networks


Institute of Electrical and Electronics Engineers



Acceptance date


Copyright date


Publisher version