CD-TVD: Contrastive Diffusion for 3D Super-Resolution with Scarce High-Resolution Time-Varying Data

2025-08-11

AI Summary

High-resolution, time-varying 3D scientific simulation data are scarce and expensive to generate, hindering effective super-resolution (SR) modeling. Method: We propose a contrastive learning–enhanced diffusion-based 3D SR framework. It integrates a contrastive encoder with a local-attention diffusion module; the encoder is pre-trained on historical low-quality data to capture degradation priors, while only one high-resolution time step is required for fine-tuning. A two-stage training strategy ensures both global consistency and local detail fidelity. Results: Experiments on fluid and atmospheric simulation datasets demonstrate that our method significantly outperforms state-of-the-art approaches under extremely limited high-resolution supervision. It achieves marked improvements in fine-grained structural reconstruction accuracy, drastically reduces reliance on large-scale, high-resolution ground-truth annotations, and enhances the practicality and scalability of scientific data augmentation.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Sampling/Simulation-based Search
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Web Mining and Content Analysis: Web data generation and simulation

Abstract

Large-scale scientific simulations require significant resources to generate high-resolution time-varying data (TVD). While super-resolution is an efficient post-processing strategy to reduce costs, existing methods rely on a large amount of HR training data, limiting their applicability to diverse simulation scenarios. To address this constraint, we proposed CD-TVD, a novel framework that combines contrastive learning and an improved diffusion-based super-resolution model to achieve accurate 3D super-resolution from limited time-step high-resolution data. During pre-training on historical simulation data, the contrastive encoder and diffusion superresolution modules learn degradation patterns and detailed features of high-resolution and low-resolution samples. In the training phase, the improved diffusion model with a local attention mechanism is fine-tuned using only one newly generated high-resolution timestep, leveraging the degradation knowledge learned by the encoder. This design minimizes the reliance on large-scale high-resolution datasets while maintaining the capability to recover fine-grained details. Experimental results on fluid and atmospheric simulation datasets confirm that CD-TVD delivers accurate and resource-efficient 3D super-resolution, marking a significant advancement in data augmentation for large-scale scientific simulations. The code is available at https://github.com/Xin-Gao-private/CD-TVD.

Problem

• Achieving 3D super-resolution with scarce high-resolution time-varying data
• Reducing reliance on large-scale HR training data for super-resolution
• Enhancing resource-efficient super-resolution for diverse simulation scenarios

Innovation

• Combines contrastive learning and diffusion model
• Uses local attention mechanism for fine-tuning
• Learns degradation patterns from limited HR data