Home /  About Me /  Team /  Projects /  Research


DECSF-Net: A multi-variable prediction method for pond aquaculture water quality based on cross-source feedback fusion
Liqiao Song, Yizhong Song, Yunchen Tian, Jianing Quan
Aquaculture International, 2025
SPRINGER
CODE

The paper shows that integrating future meteorological data into water quality prediction methods significantly improves accuracy, offering substantial practical value.

一种适用于复杂水产养殖环境的高密度鱼苗实时计数方法
Liqiao Song, Yunchen Tian, Qingfei Li, Jianing Quan
渔业现代化, 2025
中文核心

所提模型在鱼苗计数任务中参数量减少13.6%,推理速度提升15.8%,能够适应复杂背景和高密度场景的鱼苗计数需求。

Fine-grained re-identification of fish individuals based on fine-tuning of language-vision model
Jianing Quan, Can Wang, Yunchen Tian
Aquaculture
Elsevier

Comming soon! Received for production

像识别不同人类个体一样,准确地判断“谁是谁”是智能化水产养殖中的关键环节,能够广泛服务于苗种选育、病害分析、长势估计、投喂决策等诸多方面。

Benchmark dataset on feeding intensity of the pearl gentian grouper (Epinephelus fuscoguttatus♀× E. lanceolatus♂)
Haijing Qin, Yunchen Tian, Jianing Quan
Aquaculture Reports, 2025
Elsevier
CODE

A pearl gentian grouper feeding intensity assessment benchmark dataset was established for the assessment of feeding intensity and deep learning model training of recirculating aquaculture systems.

基于YOLOV8-ByteTrack 鱼苗自动计数装置的设计与试验
Rui Wang, Yunchen Tian, Jianing Quan
渔业科学进展, 2025
中文核心

研发的可适用任意品种鱼苗的自动计数算法与装置,针对3-12cm鱼苗的平均计数准确率为99.1%,算法平均帧率高达155FPS.

Machine vision-based estimation of body size and weight of pearl gentian grouper
Xueqi Cong, Yunchen Tian, Jianing Quan, Haijing Qin, Qingfei Li, Ruipeng Li
Aquaculture International, 2024
SPRINGER

The algorithm for measuring body size traits eliminates subjective factors and empirical differences in traditional methods, resulting in improved estimation accuracy.

Unambiguous pyramid cost volumes fusion for stereo matching
Qibo Chen, Baozhen Ge, Jianing Quan
IEEE TCSVT, 2024
IEEE

Unambiguous Pyramid cost volumes Fusion Network terms as UPFNet, to reduce the ambiguity between pyramid cost volumes at different scales and boost the cross-scale information flow in the stereo matching framework based on 3D convolution.

CrackViT: a unified CNN-transformer model for pixel-level crack extraction
Jianing Quan, Baozhen Ge, Min Wang
Neural Computing and Applications, 2023
SPRINGER

A systematic analysis of Transformer architectures for pixel-level crack extraction in terms of network structures and parameters, feature fusion modes, training data and strategy, and generalization ability was developed for the first time.

适用于近地面成像的自适应光学系统研究
Haiming Wang, Jianing Quan, Baozhen Ge
中国光学, 2023
Chinese Optics

为了克服近地面湍流对几十到几百米中长成像距离下光学系统成像质量的不利影响,设计了基于长焦距望远物镜和一体化自适应模块的光学成像系统。

Cross attention redistribution with contrastive learning for few shot object detection
Jianing Quan Baozhen Ge, Lei Chen
Displays, 2022
Elsevier

We focus on sufficient mining and integrating the support features conducive to generating regional proposals to improve further the stability and accuracy of the few-shot object detector.


Based on a template by Weidi Xie and Jon Barron