quantitative evaluations
Quantitative evaluations are assessments based on numerical metrics and criteria, providing objective performance measures for AI models across various tasks, facilitating comparisons and benchmarking.
- Advancing Interpretability of CLIP Representations with Concept Surrogate Model
- CREA: A Collaborative Multi-Agent Framework for Creative Image Editing and Generation
- DEXTER: Diffusion-Guided EXplanations with TExtual Reasoning for Vision Models
- DPAIL: Training Diffusion Policy for Adversarial Imitation Learning without Policy Optimization
- LuxDiT: Lighting Estimation with Video Diffusion Transformer
- MoCha: Towards Movie-Grade Talking Character Generation
- OmniVCus: Feedforward Subject-driven Video Customization with Multimodal Control Conditions
- ViewCraft3D: High-fidelity and View-Consistent 3D Vector Graphics Synthesis
- Wukong's 72 Transformations: High-fidelity Textured 3D Morphing via Flow Models
- ZeroPatcher: Training-free Sampler for Video Inpainting and Editing