The official implementation for the ECCV 2022 paper Joint Feature Learning and Relation Modeling for Tracking: A One-Stream Framework.
[Models][Raw Results][Training logs]
论文来自以上的内容Option1: Use the Anaconda (CUDA 10.2)
conda create -n ostrack python=3.8
conda activate ostrack
bash install.sh
Option2: Use the Anaconda (CUDA 11.3)
conda env create -f ostrack_cuda113_env.yaml
Option3: Use the docker file(可选)
We provide the full docker file here.
确保在./output/checkpoints/train/ostrack文件夹下,有相应的参数文件
比如参数配置为vitb_384_mae_ce_32x4_ep300,
则有 ./output/checkpoints/train/ostrackvitb_384_mae_ce_32x4_ep300/OSTrack_ep0300.pth.tar
本项目默认为vitb_384_mae_ce_32x4_ep300参数配置,OSTrack_ep0300.pth.tar
-- output
-- checkpoints
-- train
-- ostrackvitb_384_mae_ce_32x4_ep300
OSTrack_ep0300.pth.tar
本工具每次只能检测一次,确保文件夹模式如下 比如,
-- data
1.jpg
2.jpg
3.jpg
...
python detection_anno.py --imagedir 数据文件夹 --class_name 类名
比如
python detection_anno.py --imagedir /home/lei/pj2/data/cam13 --class_name jizhan
1、首次
拖动进度条选择画面,按下r画框,再按回车确认物体,然后自动推导
2、修改
选择任意画面,按r后修改,只能修改user到user之间的画面。
3、播放
黄框为确认的物体,没有遮掩的
蓝框为不确认物体,有遮掩的 最外层有红框的话,则为用户标记
标签的选择,由最底下的滑块决定
选择任意画面,按回车播放
在data文件夹下,可以发现有label/$CLASS_NAME.txt的结果
OSTrack is a simple, neat, high-performance one-stream tracking framework for joint feature learning and relational modeling based on self-attention operators. Without any additional temporal information, OSTrack achieves SOTA performance on multiple benchmarks. OSTrack can serve as a strong baseline for further research.
Tracker | GOT-10K (AO) | LaSOT (AUC) | TrackingNet (AUC) | UAV123(AUC) |
---|---|---|---|---|
OSTrack-384 | 73.7 | 71.1 | 83.9 | 70.7 |
OSTrack-256 | 71.0 | 69.1 | 83.1 | 68.3 |