代码拉取完成,页面将自动刷新
paper link: http://arxiv.org/abs/2308.12894
conda create -n ms2.0.0 python=3.7
conda activate ms2.0.0
export MS_VERSION=2.0.0
pip install -r requirements.txt
ckpt url: https://pan.baidu.com/s/1BWJi5jN5oInNCcw6K_uPag?pwd=noah
Running using SwinT-Base, and ADEChallengeData2016
python main.py ./ckpt/sdt_swin-base-4-12_640x640_ade20k_1236length_ms_total-ms.py ./ckpt/ms_convert.ckpt ./ADEChallengeData2016 --eval mIoU mFscore mDice
aAcc | mIoU | mAcc | mFscore | mPrecision | mRecall | mDice |
---|---|---|---|---|---|---|
84.57 | 54.17 | 65.46 | 67.68 | 72.03 | 65.46 | 67.68 |
@inproceedings{liu2023boosting,
title={Boosting Semantic Segmentation from the Perspective of Explicit Class Embeddings},
author={Liu, Yuhe and Liu, Chuanjian and Han, Kai and Tang, Quan and Qin, Zengchang},
booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision},
pages={821--831},
year={2023}
}
此处可能存在不合适展示的内容,页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容,可点击提交进行申诉,我们将尽快为您处理。