A5114833557- Pharmaceutical and Antibiotic Environmental Impacts
- Image and Signal Denoising Methods
- Advanced Image Fusion Techniques
- Image Enhancement Techniques
- Adversarial Robustness in Machine Learning
- Aquatic Ecosystems and Phytoplankton Dynamics
- Antimicrobial agents and applications
- Microbial Community Ecology and Physiology
- Antibiotic Use and Resistance
- Water Treatment and Disinfection
- Antibiotic Resistance in Bacteria
Zhejiang University
2018-2024
The transmission routes for antibiotic resistance genes (ARGs) and microbiota between humans water environments is poorly characterized. Here, we used high-throughput qPCR analyses 16S rRNA gene sequencing to examine the occurrence abundance of in both healthy associated from a Chinese village. Humans carried most diverse assemblage ARGs, with 234 different ARGs being detected. total feces, on skin, effluent domestic sewage treatment systems were approximately 23, 2, 7 times higher than...
Few-Shot Instance Segmentation (FSIS) requires detecting and segmenting novel classes with limited support examples. Existing methods based on Region Proposal Networks (RPNs) face two issues: 1) overfitting suppresses class objects 2) dual-branch models require complex spatial correlation strategies to prevent information loss when generating prototypes. We introduce a unified framework, Reference Twice (RefT), exploit the relationship between query features for FSIS related tasks. Our three...
Image harmonization aims to solve the visual inconsistency problem in composited images by adaptively adjusting foreground pixels with background as references. Existing methods employ local color transformation or region matching between and background, which neglects powerful proximity prior independently distinguishes fore-/back-ground a whole part for harmonization. As result, they still show limited performance across varied objects scenes. To address this issue, we propose novel...
Image harmonization aims to solve the visual inconsistency problem in composited images by adaptively adjusting foreground pixels with background as references. Existing methods employ local color transformation or region matching between and background, which neglects powerful proximity prior independently distinguishes fore-/back-ground a whole part for harmonization. As result, they still show limited performance across varied objects scenes. To address this issue, we propose novel...