# Motion Degeneracy in Self-supervised Learning of Elevation Angle Estimation for 2D Forward-Looking Sonar

arXiv: https://arxiv.org/abs/2307.16160 · HTML: https://mlsanity.com/paper/2307.16160/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 10 of its 29 references that are in the map.
- First posted on arXiv 30 Jul 2023.

## Builds on (most cited first, 10 of 10)

- [Depth Map Prediction from a Single Image using a Multi-Scale Deep Network](https://mlsanity.com/paper/1406.2283/) (2014)
- [Unsupervised Learning of Depth and Ego-Motion from Video](https://mlsanity.com/paper/1704.07813/) (2017)
- [Unsupervised CNN for Single View Depth Estimation: Geometry to the Rescue](https://mlsanity.com/paper/1603.04992/) (2016)
- [A survey of loss functions for semantic segmentation](https://mlsanity.com/paper/2006.14822/) (2020)
- [Learning Depth from Single Monocular Images Using Deep Convolutional Neural Fields](https://mlsanity.com/paper/1502.07411/) (2015)
- [How do neural networks see depth in single images?](https://mlsanity.com/paper/1905.07005/) (2019)
- [Auto-Rectify Network for Unsupervised Indoor Depth Estimation](https://mlsanity.com/paper/2006.02708/) (2020)
- [Neural Implicit Surface Reconstruction using Imaging Sonar](https://mlsanity.com/paper/2209.08221/) (2022)
- [Fusing Concurrent Orthogonal Wide-aperture Sonar Images for Dense Underwater 3D Reconstruction](https://mlsanity.com/paper/2007.10407/) (2020)
- [Learning Pseudo Front Depth for 2D Forward-Looking Sonar-based Multi-view Stereo](https://mlsanity.com/paper/2208.00233/) (2022)

## Abstract (arXiv)

2D forward-looking sonar is a crucial sensor for underwater robotic perception. A well-known problem in this field is estimating missing information in the elevation direction during sonar imaging. There are demands to estimate 3D information per image for 3D mapping and robot navigation during fly-through missions. Recent learning-based methods have demonstrated their strengths, but there are still drawbacks. Supervised learning methods have achieved high-quality results but may require further efforts to acquire 3D ground-truth labels. The existing self-supervised method requires pretraining using synthetic images with 3D supervision. This study aims to realize stable self-supervised learning of elevation angle estimation without pretraining using synthetic images. Failures during self-supervised learning may be caused by motion degeneracy problems. We first analyze the motion field of 2D forward-looking sonar, which is related to the main supervision signal. We utilize a modern learning framework and prove that if the training dataset is built with effective motions, the network can be trained in a self-supervised manner without the knowledge of synthetic data. Both simulation and real experiments validate the proposed method.