MIRO: MultI-Reward cOnditioned pretraining improves T2I quality and efficiency Paper β’ 2510.25897 β’ Published Oct 29, 2025 β’ 16
How far can we go with ImageNet for Text-to-Image generation? Paper β’ 2502.21318 β’ Published Feb 28, 2025 β’ 26
Around the World in 80 Timesteps: A Generative Approach to Global Visual Geolocation Paper β’ 2412.06781 β’ Published Dec 9, 2024 β’ 23
Don't drop your samples! Coherence-aware training benefits Conditional diffusion Paper β’ 2405.20324 β’ Published May 30, 2024
Analysis of Classifier-Free Guidance Weight Schedulers Paper β’ 2404.13040 β’ Published Apr 19, 2024
OpenStreetView-5M: The Many Roads to Global Visual Geolocation Paper β’ 2404.18873 β’ Published Apr 29, 2024
E.T. the Exceptional Trajectories: Text-to-camera-trajectory generation with character awareness Paper β’ 2407.01516 β’ Published Jul 1, 2024