Image Translation
Image translation uses artificial intelligence to transform images from one domain to another, aiming to preserve relevant information while altering style, modality, or resolution. Current research emphasizes developing efficient and controllable methods, employing architectures like generative adversarial networks (GANs) and diffusion models, often incorporating attention mechanisms and transformer networks for improved performance and handling of complex relationships between image domains. This field is significant for applications ranging from medical imaging (e.g., synthesizing missing MRI scans) and robotics (e.g., improving tactile sensor data) to creative applications like style transfer and 360° panorama manipulation, ultimately advancing both scientific understanding and practical capabilities across diverse domains.
Papers
Rethinking Perceptual Metrics for Medical Image Translation
Nicholas Konz, Yuwen Chen, Hanxue Gu, Haoyu Dong, Maciej A. Mazurowski
Lost in Translation: Modern Neural Networks Still Struggle With Small Realistic Image Transformations
Ofir Shifman, Yair Weiss
Implicit Multi-Spectral Transformer: An Lightweight and Effective Visible to Infrared Image Translation Model
Yijia Chen, Pinghua Chen, Xiangxin Zhou, Yingtie Lei, Ziyang Zhou, Mingxian Li