Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

regdb test result #25

Closed
JUN-KAI-FAN opened this issue Jun 16, 2024 · 6 comments
Closed

regdb test result #25

JUN-KAI-FAN opened this issue Jun 16, 2024 · 6 comments

Comments

@JUN-KAI-FAN
Copy link

Based on the author's provided train_regdb.bash script, I trained the model and used the following command for testing:

python test.py --tvsearch True --gpu 0 --dataset regdb

However, the results significantly differ from those reported in the author's paper. Can you explain why this discrepancy might occur?

@JUN-KAI-FAN
Copy link
Author

螢幕擷取畫面 2024-06-16 175316

@ZYK100
Copy link
Owner

ZYK100 commented Jun 16, 2024

您好,您是不是测试反了,regdb有两种模式,测试的时候需要修改一下

@JUN-KAI-FAN
Copy link
Author

您好,您是不是测试反了,regdb有两种模式,测试的时候需要修改一下

謝謝作者回復,確實照修改完後正常,另外想問作者sysu的數據集作者是否有做甚麼改動,因為測試時發現all search模式測下來跟作者差4個百分點
螢幕擷取畫面 2024-06-17 163027

@ZYK100
Copy link
Owner

ZYK100 commented Jun 17, 2024

sysu数据集没有改动,你需要按照我们的设置进行训练和测试,这个代码经过很多人的测试和实验了,表明是没问题的

@JUN-KAI-FAN
Copy link
Author

sysu数据集没有改动,你需要按照我们的设置进行训练和测试,这个代码经过很多人的测试和实验了,表明是没问题的

作者您好,剛發現我的留言可能會讓您誤會我質疑您訓練數據有偷吃步,這點跟作者誠懇的道歉,本人絕對相信作者的實驗數據是完整無誤的,也相信是沒問題的,目前發現我在pre_process_sysu.py的程式與作者有落差,重新跑過作者提供的pre_process_sysu.py並重新跑訓練,其他訓練參數皆使用作者預設的設置進行訓練,等訓練完成後會更新現況,謝謝作者的回覆

@JUN-KAI-FAN
Copy link
Author

sysu数据集没有改动,你需要按照我们的设置进行训练和测试,这个代码经过很多人的测试和实验了,表明是没问题的

作者您好,修改完後結果已經跟您實驗數據差不多了。
擷取

@ZYK100 ZYK100 closed this as completed Sep 9, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants