search

RPC-Dataset / RPC-Leaderboard

RPC Dataset Leaderboard
67 stars 9 forks source link

DPNet: Syn+Render #6

Open lufficc opened 5 years ago

lufficc commented 5 years ago

Data Priming Network for Automatic Check-Out

level method cAcc mCIoU ACD mCCD mAP50 mmAP
easy Syn+Render (DPNet) 90.32% 97.87% 0.15 0.02 98.6% 83.07%
medium Syn+Render (DPNet) 80.68% 97.38% 0.32 0.03 98.07% 77.25%
hard Syn+Render (DPNet) 70.76% 97.04% 0.53 0.03 97.76% 74.95%
averaged Syn+Render (DPNet) 80.51% 97.33% 0.34 0.03 97.91% 77.04%
ShihuaiXu commented 4 years ago

Awsome work! Could you explain your method more specifically? especially how to get the syn and render images

lufficc commented 3 years ago

Please refer to : https://github.com/lufficc/DPNet/issues/5 https://github.com/lufficc/DPNet/issues/6

ShihuaiXu commented 3 years ago

Please refer to : lufficc/DPNet#5 lufficc/DPNet#6

Thanks

ShihuaiXu commented 3 years ago

您好，我是一名湖南师范大学软件工程专业的研一的学生，拜读您的《Data Priming Network for Automatic Check-Out》，很受启发，我想按照您的论文进行复现。 为此，在此求一下做完Background Removal的单个商品的mask数据集或者利用GAN合成好的数据集，如果Backgrond removal的代码能开源就更好了。

lufficc commented 3 years ago

已开源：https://isrc.iscas.ac.cn/gitlab/research/acm-mm-2019-ACO Mask 下载地址：Google Drive

ShihuaiXu commented 3 years ago

收到 非常感谢

------------------ 原始邮件 ------------------ 发件人: "RPC-Dataset/RPC-Leaderboard" <notifications@github.com>; 发送时间: 2020年11月25日(星期三) 下午3:31 收件人: "RPC-Dataset/RPC-Leaderboard"<RPC-Leaderboard@noreply.github.com>; 抄送: "Loukas"<531232693@qq.com>;"Comment"<comment@noreply.github.com>; 主题: Re: [RPC-Dataset/RPC-Leaderboard] DPNet: Syn+Render (#6)

已开源：https://isrc.iscas.ac.cn/gitlab/research/acm-mm-2019-ACO Mask 下载地址：Google Drive

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

992060356 commented 3 years ago

收到非常感谢 … ------------------ 原始邮件 ------------------ 发件人: "RPC-Dataset/RPC-Leaderboard" <notifications@github.com>; 发送时间: 2020年11月25日(星期三) 下午3:31 收件人: "RPC-Dataset/RPC-Leaderboard"<RPC-Leaderboard@noreply.github.com>; 抄送: "Loukas"<531232693@qq.com>;"Comment"<comment@noreply.github.com>; 主题: Re: [RPC-Dataset/RPC-Leaderboard] DPNet: Syn+Render (#6) 已开源：https://isrc.iscas.ac.cn/gitlab/research/acm-mm-2019-ACO Mask 下载地址：Google Drive — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

用于自动签出的数据启动网络

等级 方法 准入 微信 ACD CCD MAP50 毫米波 简单的 同步+渲染（DPNet） 90.32% 97.87% 0.15 0.02 98.6% 83.07% 中等的 同步+渲染（DPNet） 80.68% 97.38% 0.32 0.03 98.07% 77.25% 难的 同步+渲染（DPNet） 70.76% 97.04% 0.53 0.03 97.76% 74.95% 平均 同步+渲染（DPNet） 80.51% 97.33% 0.34 0.03 97.91% 77.04%

你好，syn合成以及gan渲染的图像如何快速重新标注？