Huan Liu
Papers
3
Total Citations
48
H-Index
3
About
Huan Liu is a researcher specializing in computer vision, deep learning, and precision agriculture, with a particular focus on automated fruit detection and ripeness assessment in real-world agricultural environments. Liu's work centers on developing and optimizing state-of-the-art object detection frameworks — including transformer-based architectures and lightweight neural networks — to address the practical challenges of deploying AI systems in complex farmland settings. Among Liu's most notable contributions is the adaptation of the Swin-B transformer architecture coupled with task-aligned one-stage object detection, which has garnered 37 citations since its 2024 publication, demonstrating significant uptake within the agricultural AI community. This work addresses longstanding problems of low precision and high missing rates in strawberry ripeness detection under challenging real-world conditions. Complementing this, Liu has advanced edge computing deployment strategies through a real-time lightweight YOLO11-based framework, bridging the gap between research-grade models and practical on-farm applications. Liu's research also tackles nuanced challenges such as difficulty imbalance among detection instances, employing hybrid attention mechanisms and partial convolution-based modules to improve robustness. Collectively, Liu's contributions represent meaningful advances in intelligent agricultural automation, offering scalable, efficient solutions for food production and harvest management.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3