Jinglei Tao

Jinglei Tao

Auto-generated profile, based on publicly available information

Jinglei Tao

Georgia Institute of Technology

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications

27 October 2023
Verified
Conference Paper
Open access
HLHuajian Li
Huajian Li
HZHouke Zhao
Houke Zhao
JTJinglei Tao
Jinglei Tao
YJYifeng Jiang
Yifeng Jiang
The field of atmospheric imaging has continually grappled with the complex task of accurate contrail detection, largely due to the intricate and variable nature of contrail formations within diverse atmospheric environments. Addressing this challenge, this research introduces a pioneering computational methodology by leveraging the synergies of the EfficientNet and UNet deep learning architectures. EfficientNet, characterized by its optimized depth, width, and resolution scaling, is employed as the encoder, capturing a rich hierarchy of features with remarkable granularity. This depth of representation is crucial to delineate the subtle nuances of contrails. Following this, the UNet decoder, renowned for its symmetric expansive path that ensures pixel-wise accuracy, is tasked with the reconstruction of these segmented features, preserving spatial coherence and fine details. The ensemble of these architectures results in a robust model capable of discerning even the most nuanced contrail formations. Rigorous empirical validations were conducted, and the results affirm the model’s superior performance in contrail segmentation. The technical innovations presented in this paper not only set a new benchmark for contrail detection but also provide insights that could be pivotal for future research endeavors in atmospheric imaging using deep learning.