arXiv subjects

Haiyan Wei

Publications and source records attributed to Haiyan Wei.

3 recordsLinked to original sources

arXiv2026

Multi-History-Step SDE Inversion for Image Editing with Superior Regional Awareness

In recent years, diffusion stochastic differential equation (SDE) inversion and inversion-free methods have become prevalent for training-free image editing, as they can achieve faithful reconstruction without tuning. However, existing approaches remain inefficient, exhibit limited plasticity, and struggle to accurately preserve unedited regions. To address these issues, we propose MIEdit, a training-free editing framework based on SDE inversion. MIEdit introduces a predictor-corrector multi-history-step scheme to achieve superior editing quality with fewer steps. We further mitigate heterogeneity and conflict between the multi-conditioned noise residuals and gradient terms during sampling, improving stability and editing plasticity under large edits. MIEdit also includes Inversion-Time Automatic Semantic Angle Masking (IASM); it leverages classifier-free guidance to automatically generate semantic angle masks during inversion and applies them throughout the sampling process for regional constraints, without extra user inputs. We additionally construct EditEval++ (30 fine-grained tasks, 1,000+ image-text-mask triplets) for comprehensive evaluation; experiments show that MIEdit outperforms state-of-the-art techniques. Project page: https://whywwwzzzg.github.io/MIEdit/.

cs.CV↗

arXiv2025

Patch-Wise Hypergraph Contrastive Learning with Dual Normal Distribution Weighting for Multi-Domain Stain Transfer

Virtual stain transfer leverages computer-assisted technology to transform the histochemical staining patterns of tissue samples into other staining types. However, existing methods often lose detailed pathological information due to the limitations of the cycle consistency assumption. To address this challenge, we propose STNHCL, a hypergraph-based patch-wise contrastive learning method. STNHCL captures higher-order relationships among patches through hypergraph modeling, ensuring consistent higher-order topology between input and output images. Additionally, we introduce a novel negative sample weighting strategy that leverages discriminator heatmaps to apply different weights based on the Gaussian distribution for tissue and background, thereby enhancing traditional weighting methods. Experiments demonstrate that STNHCL achieves state-of-the-art performance in the two main categories of stain transfer tasks. Furthermore, our model also performs excellently in downstream tasks. Code is available at https://github.com/Whywwwzzzg/STNHCL

cs.CV↗

arXiv2020

A computational insight of the improved nicotine binding with ACE2-SARS-CoV-2 complex with its clinical impact

Smokers being witnessed with the mild adverse clinical symptoms of SARS-CoV-2, the in-silico study is intended to explore the effect of nicotine binding to the soluble angiotensin converting enzyme II (ACE2) receptor with or without SARS-CoV-2 binding. Nicotine established a stable interaction with the conserved amino acid residues: Asp382, Gly405, His378 and Tyr385 through His401 of the soluble ACE2 that seals its interaction with the INS1. Also, nicotine binding has significantly reduced the affinity score of ACE2 with INS1 to -12.6 kcal/mol (versus -15.7 kcal/mol without nicotine) and the interface area to 1933.6 square Angstrom (versus 2057.3 square Angstrom without nicotine). Nicotine exhibited a higher binding affinity score with ACE2-SARS-CoV-2 complex with -6.33 kcal/mol (Vs -5.24 kcal/mol without SARS-CoV-2) and a lowered inhibitory contant value of 22.95 micromolar (Vs 151.69 micromolar without SARS-CoV). Eventhough ACE2 is not a potential receptor for nicotine binding in the healthy people, in COVID19 patients, it may exhibit better binding affinity with the ACE2 receptor. In overall, nicotines strong preference for ACE2-SARS-CoV-2 complex might drastically reduce the SARS-CoV-2 virulence by intervening the ACE2 conserved residues interaction with the spike (S1) protein of SARS-CoV-2.

q-bio.BM↗