image quality
A measure of how accurately an image represents the desired object or scene, often evaluated based on clarity, detail, consistency, and fidelity. In AI, especially in generative models, improved image quality is a key objective for applications like computer vision and graphics.
- Ambient Diffusion Omni: Training Good Models with Bad Data
- Balanced Conic Rectified Flow
- BurstDeflicker: A Benchmark Dataset for Flicker Removal in Dynamic Scenes
- Entropy Rectifying Guidance for Diffusion and Flow Models
- Flow-GRPO: Training Flow Matching Models via Online RL
- Fourier Token Merging: Understanding and Capitalizing Frequency Domain for Efficient Image Generation
- LEDiT: Your Length-Extrapolatable Diffusion Transformer without Positional Encoding
- MJ-Bench: Is Your Multimodal Reward Model Really a Good Judge for Text-to-Image Generation?
- On the Coexistence and Ensembling of Watermarks
- RespoDiff: Dual-Module Bottleneck Transformation for Responsible & Faithful T2I Generation
- Spike-RetinexFormer: Rethinking Low-light Image Enhancement with Spiking Neural Networks
- The Future Unmarked: Watermark Removal in AI-Generated Images via Next-Frame Prediction
- Towards a Golden Classifier-Free Guidance Path via Foresight Fixed Point Iterations
- User-Instructed Disparity-aware Defocus Control