Picture for Mingbao Lin

Mingbao Lin

TextDestroyer: A Training- and Annotation-Free Diffusion Method for Destroying Anomal Text from Images

Add code
Nov 01, 2024
Viaarxiv icon

UniVST: A Unified Framework for Training-free Localized Video Style Transfer

Add code
Oct 26, 2024
Viaarxiv icon

EasyInv: Toward Fast and Better DDIM Inversion

Add code
Aug 09, 2024
Viaarxiv icon

Advancing Multimodal Large Language Models with Quantization-Aware Scale Learning for Efficient Adaptation

Add code
Aug 07, 2024
Viaarxiv icon

Move and Act: Enhanced Object Manipulation and Background Integrity for Image Editing

Add code
Jul 25, 2024
Figure 1 for Move and Act: Enhanced Object Manipulation and Background Integrity for Image Editing
Figure 2 for Move and Act: Enhanced Object Manipulation and Background Integrity for Image Editing
Figure 3 for Move and Act: Enhanced Object Manipulation and Background Integrity for Image Editing
Figure 4 for Move and Act: Enhanced Object Manipulation and Background Integrity for Image Editing
Viaarxiv icon

AccDiffusion: An Accurate Method for Higher-Resolution Image Generation

Add code
Jul 15, 2024
Viaarxiv icon

AnySR: Realizing Image Super-Resolution as Any-Scale, Any-Resource

Add code
Jul 05, 2024
Viaarxiv icon

LLMs-as-Instructors: Learning from Errors Toward Automating Model Improvement

Add code
Jun 29, 2024
Viaarxiv icon

UIO-LLMs: Unbiased Incremental Optimization for Long-Context LLMs

Add code
Jun 26, 2024
Viaarxiv icon

UniPTS: A Unified Framework for Proficient Post-Training Sparsity

Add code
May 29, 2024
Viaarxiv icon