UTVit: A U-Shaped Segmentation Network for Underwater Images Based on TinyVit

TY - GEN

T1 - UTVit

T2 - 44th Chinese Control Conference, CCC 2025

AU - Wang, Kai

AU - Lu, Pingli

AU - Zou, Suli

AU - Hu, Hengzai

N1 - Publisher Copyright: © 2025 Technical Committee on Control Theory, Chinese Association of Automation.

PY - 2025

Y1 - 2025

N2 - This paper presents a lightweight semantic segmentation model for underwater hazardous objects based on the TinyViT backbone network, realizing a real-time semantic segmentation under the computational constraints of unmanned underwater vehicles (UUVs). To efficiently leverage feature maps from various network layers, we devise a network architecture based on U-Net named UTVit, incorporating skip connections and concatenations to integrate feature maps across different levels of the network. Furthermore, we engineered a specialized upsampling network that employs depth-wise convolutions instead of standard convolutions. This approach mitigates the coarse granularity issues often encountered during upsampling to significantly reduce the computational burden. To validate the effectiveness of our proposed model, we conduct extensive experiments on the publicly available USIS10K dataset, comparing our model with state-of-the-art (SOTA) models. The results show that UTVit achieves a slight 2.3% decrease in mIOU (mean Intersection over Union) while significantly reducing the number of parameters by 51.46M, remaining a 12.7% parameter count of the former. In the parameter-mIOU and computation-mIOU trade-off charts, UTVit exhibits superior performance, making it highly suitable for object segmentation tasks in UUV applications. This balance between efficiency and effectiveness underscores the model's potential for practical deployment in resource-constrained underwater environments.

AB - This paper presents a lightweight semantic segmentation model for underwater hazardous objects based on the TinyViT backbone network, realizing a real-time semantic segmentation under the computational constraints of unmanned underwater vehicles (UUVs). To efficiently leverage feature maps from various network layers, we devise a network architecture based on U-Net named UTVit, incorporating skip connections and concatenations to integrate feature maps across different levels of the network. Furthermore, we engineered a specialized upsampling network that employs depth-wise convolutions instead of standard convolutions. This approach mitigates the coarse granularity issues often encountered during upsampling to significantly reduce the computational burden. To validate the effectiveness of our proposed model, we conduct extensive experiments on the publicly available USIS10K dataset, comparing our model with state-of-the-art (SOTA) models. The results show that UTVit achieves a slight 2.3% decrease in mIOU (mean Intersection over Union) while significantly reducing the number of parameters by 51.46M, remaining a 12.7% parameter count of the former. In the parameter-mIOU and computation-mIOU trade-off charts, UTVit exhibits superior performance, making it highly suitable for object segmentation tasks in UUV applications. This balance between efficiency and effectiveness underscores the model's potential for practical deployment in resource-constrained underwater environments.

KW - Depth-wise Convolution

KW - Lightweight Model

KW - Underwater Semantic Segmentation

KW - Vision Transformer

UR - https://www.scopus.com/pages/publications/105020272191

U2 - 10.23919/CCC64809.2025.11179506

DO - 10.23919/CCC64809.2025.11179506

M3 - Conference contribution

AN - SCOPUS:105020272191

T3 - Chinese Control Conference, CCC

SP - 7863

EP - 7868

BT - Proceedings of the 44th Chinese Control Conference, CCC 2025

A2 - Sun, Jian

A2 - Yin, Hongpeng

PB - IEEE Computer Society

Y2 - 28 July 2025 through 30 July 2025

ER -