Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Synth-SONAR: Sonar Image Synthesis with Enhanced Diversity and Realism via Dual Diffusion Models and GPT Prompting

Purushothaman Natarajan, Kamal Basha, Athira Nambiar

TL;DR

This study proposes a new sonar image synthesis framework, Synth-SONAR leveraging diffusion models and GPT prompting, and achieves state-of-the-art results in producing high-quality synthetic sonar datasets, significantly enhancing their diversity and realism.

Abstract

Sonar image synthesis is crucial for advancing applications in underwater exploration, marine biology, and defence. Traditional methods often rely on extensive and costly data collection using sonar sensors, jeopardizing data quality and diversity. To overcome these limitations, this study proposes a new sonar image synthesis framework, Synth-SONAR leveraging diffusion models and GPT prompting. The key novelties of Synth-SONAR are threefold: First, by integrating Generative AI-based style injection techniques along with publicly available real/simulated data, thereby producing one of the largest sonar data corpus for sonar research. Second, a dual text-conditioning sonar diffusion model hierarchy synthesizes coarse and fine-grained sonar images with enhanced quality and diversity. Third, high-level (coarse) and low-level (detailed) text-based sonar generation methods leverage advanced semantic information available in visual language models (VLMs) and GPT-prompting. During inference, the method generates diverse and realistic sonar images from textual prompts, bridging the gap between textual descriptions and sonar image generation. This marks the application of GPT-prompting in sonar imagery for the first time, to the best of our knowledge. Synth-SONAR achieves state-of-the-art results in producing high-quality synthetic sonar datasets, significantly enhancing their diversity and realism.

Synth-SONAR: Sonar Image Synthesis with Enhanced Diversity and Realism via Dual Diffusion Models and GPT Prompting

TL;DR

This study proposes a new sonar image synthesis framework, Synth-SONAR leveraging diffusion models and GPT prompting, and achieves state-of-the-art results in producing high-quality synthetic sonar datasets, significantly enhancing their diversity and realism.

Abstract

Sonar image synthesis is crucial for advancing applications in underwater exploration, marine biology, and defence. Traditional methods often rely on extensive and costly data collection using sonar sensors, jeopardizing data quality and diversity. To overcome these limitations, this study proposes a new sonar image synthesis framework, Synth-SONAR leveraging diffusion models and GPT prompting. The key novelties of Synth-SONAR are threefold: First, by integrating Generative AI-based style injection techniques along with publicly available real/simulated data, thereby producing one of the largest sonar data corpus for sonar research. Second, a dual text-conditioning sonar diffusion model hierarchy synthesizes coarse and fine-grained sonar images with enhanced quality and diversity. Third, high-level (coarse) and low-level (detailed) text-based sonar generation methods leverage advanced semantic information available in visual language models (VLMs) and GPT-prompting. During inference, the method generates diverse and realistic sonar images from textual prompts, bridging the gap between textual descriptions and sonar image generation. This marks the application of GPT-prompting in sonar imagery for the first time, to the best of our knowledge. Synth-SONAR achieves state-of-the-art results in producing high-quality synthetic sonar datasets, significantly enhancing their diversity and realism.

Paper Structure

This paper contains 28 sections, 17 equations, 9 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Underwater Sonar image analysis
  4. Synthetic Data Generation via Generative AI
  5. Vision-Language Models (VLMs) and Their Application in the Sonar Domain
  6. Methodology
  7. Phase-1: Data Acquisition
  8. Style Injection on Generated Images
  9. Phase-2: Train and Generate sonar Images via DDPM and GPT-prompting
  10. Phase-3: Tune & Generalize DDPM for Fine-Grained sonar Image Generation
  11. Experimental Setup
  12. Dataset
  13. Publicly Available Sonar Images
  14. S3 Simulator Images
  15. Evaluation Metrics
  16. ...and 13 more sections

Figures (9)

  • Figure 1: Overall Architecture of the proposed "Synth-SONAR" sonar image synthesis framework.
  • Figure 2: Style Injection for sonar Image Synthesis.
  • Figure 3: Samples from publicly available sonar images
  • Figure 4: Samples from S3 Simulator Dataset
  • Figure 5: The Contents, Styles, and Stylized sonar Images (Outputs) in the style injection process
  • ...and 4 more figures