>mfw Research news

10/09/2025

>Control-Augmented Autoregressive Diffusion for Data Assimilation
https://arxiv.org/abs/2510.06637

>OBS-Diff: Accurate Pruning For Diffusion Models in One-Shot
https://arxiv.org/abs/2510.06751

>DreamOmni2: Multimodal Instruction-based Editing and Generation
https://arxiv.org/abs/2510.06679

>MV-Performer: Taming Video Diffusion Model for Faithful and Synchronized Multi-view Performer Synthesis
https://arxiv.org/abs/2510.07190

>Enhancing Concept Localization in CLIP-based Concept Bottleneck Models
https://arxiv.org/abs/2510.07115

>U-Bench: A Comprehensive Understanding of U-Net through 100-Variant Benchmarking
https://fenghetan9.github.io/ubench

>Sharpness-Aware Data Generation for Zero-shot Quantization
https://arxiv.org/abs/2510.07018

>IAR2: Improving Autoregressive Visual Generation with Semantic-Detail Associated Token Prediction
https://arxiv.org/abs/2510.06928

>SIGMA-GEN: Structure and Identity Guided Multi-subject Assembly for Image Generation
https://oindrilasaha.github.io/SIGMA-Gen

>Ming-UniVision: Joint Image Understanding and Generation with a Unified Continuous Tokenizer
https://arxiv.org/abs/2510.06590

>VUGEN: Visual Understanding priors for GENeration
https://arxiv.org/abs/2510.06529

>Efficient High-Resolution Image Editing with Hallucination-Aware Loss and Adaptive Tiling
https://arxiv.org/abs/2510.06295

>Rethinking Inter-LoRA Orthogonality in Adapter Merging: Insights from Orthogonal Monte Carlo Dropout
https://arxiv.org/abs/2510.03262

10/08/2025

>Data Factory with Minimal Human Effort Using VLMs
https://arxiv.org/abs/2510.05722

>Teleportraits: Training-Free People Insertion into Any Scene
https://arxiv.org/abs/2510.05660

>Beyond Spectral Peaks: Interpreting the Cues Behind Synthetic Image Detection
https://arxiv.org/abs/2510.05633

>Efficient Conditional Generation on Scale-based Visual Autoregressive Models
https://arxiv.org/abs/2510.05610