A deep learning-based approach for generating realistic foggy images using ControlNet, LoRA and FOHIS. This project fine-tunes diffusion models for improved fog simulation, useful in training autonomous vehicles and environmental simulations.
This simulator uses the FOHIS (Fog in High Intensity Scenarios) approach to generate varying levels of fog density, enhancing the realism of foggy conditions in simulated environments.
This project utilizes stable diffusion techniques to simulate foggy conditions in images, offering a versatile framework for generating synthetic data in controlled environments.