Imaging from Deepfake Data - Applications in Seismic Imaging

Explore the innovative applications of neural networks in imaging and inversion from sensor data and physical models in this virtual seminar talk by Laurent Demanet from MIT. Delve into the concept of generating "deepfake" data to enhance inversion techniques, with a focus on seismic imaging. Discover three key examples: bandwidth extension, data from virtual sources, and "physics swap". Gain insights into collaborative research efforts aimed at expanding the capabilities of imaging technologies through artificial intelligence.