ETETMO: An End-to-End Visual Intelligent Surveillance Framework for Robust Tracking of Moving Objects on the Airport Surface
1 Nanjing University of Aeronautics and Astronautics, 2Chinese Academy of Civil Aviation Science and Technology
π§ corresponding author.
-
2/07/2025
: We release the pre-trained models and the detect code. -
1/07/2025
: We have published some visualization experiments static and dynamic results. -
11/02/2024
: We submitted the paper. Code and pre-trained model are coming soon.
This project contains the official PyTorch implementation, pre-trained models, fine-tuning code, and detect demo for ETETMO.
-
An end-to-end intelligent visual surveillance framework,ETETMO, is proposed for robust tracking of moving objects on the airport surface.
-
ETETMO replaces the traditional manually designed ID association methods with a learnable ID classification strategy.
-
ETETMO surpasses state-of-the-art methods across five key evaluation metrics, achieving a 7.6% improvement in HOTA and an 11% improvement in IDF1.
ETETMO
βββ dataset
β βββ ASV-T2024
β β βββ annotations
β β βββ train
β β βββ val
β β βββ test
β βββ coco
β β βββ annotations
β β βββ train2017
β β βββ val2017
Compile and install the required packages
pip install timm==0.9.8 thop efficientnet_pytorch==0.7.1 einops grad-cam==1.4.8 dill==0.3.6 albumentations==1.3.1 pytorch_wavelets==1.3.0 tidecv PyWavelets -i https://pypi.tuna.tsinghua.edu.cn/simple
pip install -U openmim
mim install mmengine
mim install "mmcv>=2.0.0"
This project has referenced some excellent open-sourced repos (Detectron2, detrex, RT-DETR, ultralytics). Thanks for their wonderful works and contributions to the community.