TaylorShift: Shifting the Complexity of Self-Attention from Squared to Linear (and Back) using Taylor-Softmax Paper • 2403.02920 • Published Mar 5, 2024
Zoomed In, Diffused Out: Towards Local Degradation-Aware Multi-Diffusion for Extreme Image Super-Resolution Paper • 2411.12072 • Published Nov 18, 2024
ForAug: Recombining Foregrounds and Backgrounds to Improve Vision Transformer Training with Bias Mitigation Paper • 2503.09399 • Published 14 days ago