Semantic Segmentation of Underwater Imagery: Dataset and Benchmark (2020-04-02T00:00:00.000000Z)

TL;DR

The first large-scale dataset for semantic Segmentation of Underwater IMagery (SUIM) is presented and SUIM-Net, a fully-convolutional deep residual model that balances the trade-off between performance and computational efficiency is presented.

Abstract

In this paper, we present the first large-scale dataset for semantic Segmentation of Underwater IMagery (SUIM). It contains over 1500 images with pixel annotations for eight object categories: fish (vertebrates), reefs (invertebrates), aquatic plants, wrecks/ruins, human divers, robots, and sea-floor. The images have been rigorously collected during oceanic explorations and human-robot collaborative experiments, and annotated by human participants. We also present a comprehensive benchmark evaluation of several state-of-the-art semantic segmentation approaches based on standard performance metrics. Additionally, we present SUIM-Net, a fully-convolutional deep residual model that balances the trade-off between performance and computational efficiency. It offers competitive performance while ensuring fast end-to-end inference, which is essential for its use in the autonomy pipeline by visually-guided underwater robots. In particular, we demonstrate its usability benefits for visual servoing, saliency prediction, and detailed scene understanding. With a variety of use cases, the proposed model and benchmark dataset open up promising opportunities for future research in underwater robot vision.

Authors

M. Islam

4 papers

Chelsey Edge

1 papers

Yuyang Xiao

1 papers

TL;DR

Abstract

Authors

References54 items

Simultaneous Enhancement and Super-Resolution of Underwater Imagery for Improved Visual Perception

CoralSeg: Learning coral segmentation from sparse annotations

Underwater Image Super-Resolution using Deep Residual Multipliers

Improving Underwater Obstacle Detection using Semantic Image Segmentation

Fast Underwater Image Enhancement for Improved Visual Perception

Toward a Generic Diver-Following Algorithm: Balancing Robustness and Efficiency in Deep Visual Detection

Salient Object Detection Driven by Fixation Prediction

Generalized Cross Entropy Loss for Training Deep Neural Networks with Noisy Labels

Understanding human motion and gestures for underwater human–robot collaboration

Underwater multi-robot convoying using visual tracking by detection

Mixed-domain biological motion tracking for underwater human-robot interaction

Multi-level Attention Networks for Visual Question Answering

Rethinking Atrous Convolution for Semantic Image Segmentation

Underwater image segmentation with co-saliency detection and local statistical active contour model

A Review on Deep Learning Techniques Applied to Semantic Segmentation

MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications

Pyramid Scene Parsing Network

MapReduce-based fast fuzzy c-means algorithm for large-scale underwater image segmentation

A Multi-scale CNN for Affordance Segmentation in RGB Images

Robotic Visual Tracking of Relevant Cues in Underwater Environments with Poor Visibility Conditions

DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs

A Benchmark Dataset and Evaluation Methodology for Video Object Segmentation

Underwater image feature extraction and matching based on visual saliency detection

Recurrent Mixture Density Network for Spatiotemporal Visual Attention

Deep Residual Learning for Image Recognition

Multi-Scale Context Aggregation by Dilated Convolutions

ReSeg: A Recurrent Neural Network-Based Model for Semantic Segmentation

ParseNet: Looking Wider to See Better

SegNet: A Deep Convolutional Encoder-Decoder Architecture for Robust Semantic Pixel-Wise Labelling

U-Net: Convolutional Networks for Biomedical Image Segmentation

Automated Fish Detection in Underwater Images Using Shape‐Based Level Sets

Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift

Adam: A Method for Stochastic Optimization

Fully convolutional networks for semantic segmentation

Visual summaries for low-bandwidth semantic mapping with autonomous underwater vehicles

Going deeper with convolutions

Very Deep Convolutional Networks for Large-Scale Image Recognition

Learning and Transferring Mid-level Image Representations Using Convolutional Neural Networks

Microsoft COCO: Common Objects in Context

Autonomous adaptive exploration using realtime online spatiotemporal topic modeling

Multi-domain monitoring of marine environments using a heterogeneous robot team

Automated annotation of coral reef survey images

Automatic fish segmentation via double local thresholding for trawl-based underwater camera systems

Robotic tools for deep water archaeology: Surveying an ancient shipwreck with an autonomous underwater vehicle

Rectified Linear Units Improve Restricted Boltzmann Machines

Deconvolutional networks

A Vision-Based Control and Interaction Framework for a Legged Underwater Robot

Underwater Scene Segmentation by Deep Neural Network

VIAME Datasets and Challenges

Non linear Image segmentation using fuzzy c means clustering method with thresholding for underwater images

Objective Assessment of Nonlinear Segmentation Approaches to Gray Level Underwater Images

“Marine Species Identiﬁcation Portal,”

“The Ocean Animal Encyclopedia,”

This Paper Is Included in the Proceedings of the 12th Usenix Symposium on Operating Systems Design and Implementation (osdi '16). Tensorflow: a System for Large-scale Machine Learning Tensorflow: a System for Large-scale Machine Learning

Field of Study

Journal Information

Name

Page

Venue Information

Name

Type

URL

Alternate Names