Paper-Conference.

• 2026 When Pretty Isn't Useful: Investigating Why Modern Text-to-Image Models Fail as Reliable Training Data Generators
• 2025 HyperCore: Coreset Selection under Noise via Hypersphere Models
• 2025 When 512×512 is not Enough: Local Degradation-Aware Multi-Diffusion for Extreme Image Super-Resolution
• 2025 Which Transformer to Favor: A Comparative Analysis of Efficiency in Vision Transformers
• 2025 A Study in Dataset Distillation for Image Super-Resolution
• 2024 TaylorShift: Shifting the Complexity of Self-Attention from Squared to Linear (and Back) using Taylor-Softmax
• 2024 A Low-Resolution Image is Worth 1x1 Words: Enabling Fine Image Super-Resolution with Transformers and TaylorShift