DUViN: Diffusion-Based Underwater Visual Navigation via Knowledge-Transferred Depth Features

2025/09/03 by Jinghe Yang, Yang, Jinghe, Minh-Quan Le +5 · 1 citation
Computer Science · Engineering · #Advanced Vision and Imaging #Channel (broadcasting) #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Feature (linguistics) #Generalization #Image Enhancement Techniques #Robotics (cs.RO) #Task (project management) #Terrain #Underwater #Underwater Vehicles and Communication Systems

paper · pdf · doi:10.48550/arxiv.2509.02983

published in arXiv (Cornell University) (Cornell University)

openalex publication_date 2025/09/03 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05

Abstract

Autonomous underwater navigation remains a challenging problem due to limited sensing capabilities and the difficulty of constructing accurate maps in underwater environments. In this paper, we propose a Diffusion-based Underwater Visual Navigation policy via knowledge-transferred depth features, named DUViN, which enables vision-based end-to-end 4-DoF motion control for underwater vehicles in unknown environments. DUViN guides the vehicle to avoid obstacles and maintain a safe and perception awareness altitude relative to the terrain without relying on pre-built maps. To address the difficulty of collecting large-scale underwater navigation datasets, we propose a method that ensures robust generalization under domain shifts from in-air to underwater environments by leveraging depth features and introducing a novel model transfer strategy. Specifically, our training framework consists of two phases: we first train the diffusion-based visual navigation policy on in-air datasets using a pre-trained depth feature extractor. Secondly, we retrain the extractor on an underwater depth estimation task and integrate the adapted extractor into the trained navigation policy from the first step. Experiments in both simulated and real-world underwater environments demonstrate the effectiveness and generalization of our approach. The experimental videos are available at https://www.youtube.com/playlist?list=PLqt2s-RyCf1gfXJgFzKjmwIqYhrP4I-7Y.

Citations

• Knowledge Distillation for Underwater Feature Extraction and Matching via GAN-synthesized Images
• Depth Anything V2
• Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data
• FAPP: Fast and Adaptive Perception and Planning for UAVs in Dynamic Cluttered Environments
• NoMaD: Goal Masked Diffusion Policies for Navigation and Exploration
• ViNT: A Foundation Model for Visual Navigation
• DINOv2: Learning Robust Visual Features without Supervision
• Diffusion Policy: Visuomotor Policy Learning via Action Diffusion
• Diffusion policy: Visuomotor policy learning via action diffusion
• Navigating to Objects in the Real World
• Self-Supervised Monocular Depth Underwater
• UDepth: Fast Monocular Depth Estimation for Visually-guided Underwater Robots
• AquaVis: A Perception-Aware Autonomous Navigation Framework for Underwater Vehicles
• Rapid Exploration for Open-World Navigation with Latent Goal Models
• Vision Transformers for Dense Prediction
• Visual Navigation in Real-World Indoor Environments Using End-to-End Deep Reinforcement Learning
• Vision-Based Goal-Conditioned Policies for Underwater Navigation in the Presence of Obstacles
• Denoising Diffusion Probabilistic Models
• Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfer
• Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-Shot Cross-Dataset Transfer
• Unsupervised Single Image Underwater Depth Estimation
• FIESTA: Fast Incremental Euclidean Distance Fields for Online Motion Planning of Aerial Robots
• Underwater Single Image Color Restoration Using Haze-Lines and a New Quantitative Dataset
• Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor
• End-to-end Driving via Conditional Imitation Learning
• Self-supervised Deep Reinforcement Learning with Generalized Computation Graphs for Robot Navigation
• FiLM: Visual Reasoning with a General Conditioning Layer
• Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization
• Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World
• End to End Learning for Self-Driving Cars
• Continuous control with deep reinforcement learning
• Depth Map Prediction from a Single Image using a Multi-Scale Deep Network
• Handbook of Marine Craft Hydrodynamics and Motion Control
• A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning

Cited by

• DREAM: Domain-aware Reasoning for Efficient Autonomous Underwater Monitoring

Related

• Atlantis: Enabling Underwater Depth Estimation with Stable Diffusion
• UIVNAV: Underwater Information-driven Vision-based Navigation via Imitation Learning
• Vision-Based Goal-Conditioned Policies for Underwater Navigation in the Presence of Obstacles
• UnderwaterVLA: Dual-brain Vision-Language-Action architecture for Autonomous Underwater Navigation
• Ocean Diviner: A Diffusion-Augmented Reinforcement Learning Framework for AUV Robust Control in Underwater Tasks