Answers are grounded in this paper and cite their sources.
The advancement of intelligent marine systems requires realistic and data-driven virtual environments for effective analysis and game-based simulation. This research proposes an integrated method for marine resource simulation and game scene construction based on the Internet of Things (IoT) and deep learning (DL). The Marine IoT Dataset with Environmental and Image consists of 6745 records integrating oceanographic parameters with corresponding marine scene image identifiers, including temperature, salinity, turbidity, current velocity, and biological observations. These multi-source data undergo preprocessing steps such as median filtering to remove noise and normalization to ensure consistency and completeness. Feature extraction is performed using Convolutional Neural Networks (CNNs) to capture spatial patterns and semantic representations of marine scenes. The proposed method employs a Satin Bowerbird Optimizer-driven Conditional Generative Adversarial Network (SBO-ConGAN) to generate high-resolution virtual marine environments, including seabed structures, water textures, and marine species distributions, while preserving environmental semantics. The generator produces realistic scenes conditioned on environmental parameters, and the discriminator enhances visual fidelity by distinguishing synthetic outputs from real data. The generated outputs are integrated into a real-time rendering engine for interactive game scene construction, incorporating dynamic environmental effects and adaptive marine object behaviors. Experimental results, obtained through Python implementation, demonstrate that the proposed SBO-ConGAN achieves an Inception Score (IS) of 9.57 and a Fréchet Inception Distance (FID) of 17.4, indicating higher realism and consistency. Overall, this proposed method provides a scalable, cost-effective solution for intelligent marine visualization and immersive game development, supporting applications in marine research, education, and interactive simulation systems. IoT-enabled marine data integration for realistic virtual environment simulation. CNN-based feature extraction captures semantic patterns in marine scenes. SBO-ConGAN generates high-fidelity marine environments for interactive games. It supports scalable, immersive marine visualization and research applications.
No takes yet. Share an insight, caveat, or question.
Chang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: