trained-model

[happyStudyOk]

Hi,Thank you for your work. I did not have access when downloading the trained-model. Could you please resend it to me? Thank you!

[stevenlsw]

Hi, I update the pre-trained model link, you could download follow README.md.

[happyStudyOk]

Thank you very much！May I ask which versions of cuda and pytorch you are using？

------------------ 原始邮件 ------------------ 发件人: "stevenlsw/Semi-Hand-Object" @.>; 发送时间: 2024年3月30日(星期六) 凌晨0:31 @.>; @.**@.>; 主题: Re: [stevenlsw/Semi-Hand-Object] trained-model (Issue #16)

Hi, I update the pre-trained model link, you could download follow README.md.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

[stevenlsw]

Hi, i think there are no specific requirements for the cuda and pytorch, pytorch version before 2.0 should work. Please let me know if this doesn't work.

[happyStudyOk]

Thank you very much. I also encountered a question about how to obtain evaluation metrics after testing. I found that I cannot use them when entering Codelab. Thank you for your answer and help.

------------------ 原始邮件 ------------------ 发件人: "stevenlsw/Semi-Hand-Object" @.>; 发送时间: 2024年4月13日(星期六) 中午1:52 @.>; @.**@.>; 主题: Re: [stevenlsw/Semi-Hand-Object] trained-model (Issue #16)

Hi, i think there are no specific requirements for the cuda and pytorch, pytorch version before 2.0 should work. Please let me know if this doesn't work.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

[happyStudyOk]

Thank you very much. This is the result I ran out of, but it's not satisfactory. Can you help me see why?

Thank you！ ------------------ 原始邮件 ------------------ 发件人: "stevenlsw/Semi-Hand-Object" @.>; 发送时间: 2024年4月13日(星期六) 中午1:52 @.>; @.**@.>; 主题: Re: [stevenlsw/Semi-Hand-Object] trained-model (Issue #16)

Hi, i think there are no specific requirements for the cuda and pytorch, pytorch version before 2.0 should work. Please let me know if this doesn't work.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

[stevenlsw]

Hi, I could not see the results attached, you should obtain the evaluation from Codelab, and evaluations are computed follow here

[happyStudyOk]

Thank you! I have obtained the results, but there is a big gap between the results and yours. I don't know why, but the results are in my attachment. Thank you very much for helping me check the reasons. ------------------ 原始邮件 ------------------ 发件人: "stevenlsw/Semi-Hand-Object" @.>; 发送时间: 2024年5月8日(星期三) 下午2:41 @.>; @.**@.>; 主题: Re: [stevenlsw/Semi-Hand-Object] trained-model (Issue #16)

Hi, I could not see the results attached, you should obtain the evaluation from Codelab, and evaluations are computed follow here

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

[stevenlsw]

Hi, I couldn't see the attachments. Are you use the trained model?

[happyStudyOk]

Thank you for your reply! I used a training model that I ran on my own, but there is a big gap between the results and yours.This is the result. Collecting open3d-python   Downloading https://files.pythonhosted.org/packages/5f/5c/a86082dc5efc3d22585e8aa22f9840667d9faa5e727b47c43137090caed4/open3d_python-0.7.0.0-cp27-cp27mu-manylinux1_x86_64.whl (3.7MB) Requirement already satisfied: numpy in /opt/conda/lib/python2.7/site-packages (from open3d-python) Requirement already satisfied: notebook in /opt/conda/lib/python2.7/site-packages (from open3d-python) Collecting widgetsnbextension (from open3d-python)   Downloading https://files.pythonhosted.org/packages/9c/a0/ba2634cd75b7d7f8f9aeb38edf854cd6c9877ec064013a62630b4541b88f/widgetsnbextension-3.6.6-py2.py3-none-any.whl (1.6MB) Requirement already satisfied: ipywidgets in /opt/conda/lib/python2.7/site-packages (from open3d-python) Installing collected packages: widgetsnbextension, open3d-python Successfully installed open3d-python-0.7.0.0 widgetsnbextension-3.6.6 Loading predictions from /tmp/codalab/tmpuf9m6O/run/input/res/pred.json Evaluation 3D KP results: auc=0.316, mean_kp3d_avg=4.91 cm Evaluation 3D KP PROCRUSTES ALIGNED results: auc=0.757, mean_kp3d_avg=1.22 cm Evaluation 3D KP SCALE-TRANSLATION ALIGNED results: auc=0.331, mean_kp3d_avg=4.85 cm

Evaluation 3D MESH results: auc=0.325, mean_kp3d_avg=4.75 cm Evaluation 3D MESH ALIGNED results: auc=0.765, mean_kp3d_avg=1.18 cm

F-scores @. = 0.152 @. = 0.447 @. = 0.513 @. = 0.916 Scores written to: /tmp/codalab/tmpuf9m6O/run/output/scores.txt Evaluation complete.

------------------ 原始邮件 ------------------ 发件人: "stevenlsw/Semi-Hand-Object" @.>; 发送时间: 2024年5月13日(星期一) 下午4:51 @.>; @.**@.>; 主题: Re: [stevenlsw/Semi-Hand-Object] trained-model (Issue #16)

Hi, I couldn't see the attachments. Are you use the trained model?

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

[happyStudyOk]

Hello, I would like to ask why I tested using the training model you provided and the results were not good? How to solve it? Sincere thanks!、 Trying to locate the prediction file automatically ... Found file "/tmp/codalab/tmpJQSQci/run/input/res/pred{}.json" Loading predictions from /tmp/codalab/tmpJQSQci/run/input/res/pred{}.json Evaluation 3D KP results: auc=0.086, mean_kp3d_avg=12.73 cm Evaluation 3D KP PROCRUSTES ALIGNED results: auc=0.724, mean_kp3d_avg=1.38 cm Evaluation 3D KP SCALE-TRANSLATION ALIGNED results: auc=0.088, mean_kp3d_avg=12.40 cm

Evaluation 3D MESH results: auc=0.049, mean_kp3d_avg=12.40 cm Evaluation 3D MESH ALIGNED results: auc=0.731, mean_kp3d_avg=1.35 cm

F-scores @. = 0.037 @. = 0.362 @. = 0.183 @. = 0.898 Scores written to: /tmp/codalab/tmpJQSQci/run/output/scores.txt Evaluation complete.