>mfw Research news

07/10/2025

>PromptTea: Let Prompts Tell TeaCache the Optimal Threshold
https://arxiv.org/abs/2507.06739

>Enhancing Diffusion Model Stability for Image Restoration via Gradient Management
https://arxiv.org/abs/2507.06656

>MS-DPPs: Multi-Source Determinantal Point Processes for Contextual Diversity Refinement of Composite Attributes in Text to Image Retrieval
https://arxiv.org/abs/2507.06654

>Towards Multimodal Understanding via Stable Diffusion as a Task-Aware Feature Extractor
https://vatsalag99.github.io/mustafar

>Concept-TRAK: Understanding how diffusion models learn concepts through concept-level attribution
https://arxiv.org/abs/2507.06547

>Spatial-Temporal Graph Mamba for Music-Guided Dance Video Synthesis
https://arxiv.org/abs/2507.06689

>Vision-Language-Vision Auto-Encoder: Scalable Knowledge Distillation from Diffusion Models
https://arxiv.org/abs/2507.07104

>Democratizing High-Fidelity Co-Speech Gesture Video Generation
https://arxiv.org/abs/2507.06812

>Denoising Multi-Beta VAE: Representation Learning for Disentanglement and Generation
https://arxiv.org/abs/2507.06613

>Evaluating Attribute Confusion in Fashion Text-to-Image Generation
https://intelligolabs.github.io/L-VQAScore

>Physics-Grounded Motion Forecasting via Equation Discovery for Trajectory-Guided Image-to-Video Generation
https://arxiv.org/abs/2507.06830

>3D-Generalist: Self-Improving Vision-Language-Action Models for Crafting 3D Worlds
https://ai.stanford.edu/~sunfanyun/3d-generalist

>Concept Unlearning by Modeling Key Steps of Diffusion Process
https://arxiv.org/abs/2507.06526

>FIFA: Unified Faithfulness Evaluation Framework for Text-to-Video and Video-to-Text Generation
https://arxiv.org/abs/2507.06523

>DynamicID: Zero-Shot Multi-ID Image Personalization with Flexible Facial Editability
https://arxiv.org/abs/2503.06505