MODNet icon indicating copy to clipboard operation
MODNet copied to clipboard

A Trimap-Free Portrait Matting Solution in Real Time [AAAI 2022]

Results 72 MODNet issues
Sort by recently updated
recently updated
newest added

运行MODNet离线demo——WebCam,在调用摄像头进行人像抠图,延迟比较严重。有没有什么解决方法

Can someone kindly provide a code of dataloader, thanks?

![image](https://user-images.githubusercontent.com/31236750/175809016-3644a42d-fc53-477b-a209-36daed33bc6d.png) 效果很差,请问是可以调整什么阈值什么的吗

Hey, I read your paper and its awesome!, I trained it on 512x512 resolution, it is really good, but when i trained the same data on higher resolution like 1024x1024...

您好,看到您在一个issue下的评论说到商用版本用到了检测分支,是在原模型中加入一个新的分支得到人像的检测框,进一步使得人像抠图更加准确吗?

Hello Everyone, First of all this project is inspiring, I would like to thank the amazing developers YOU ROCKS! 💙 Just to be clear, English is not my native language...

Thank you for your great research & repository, and congrats on MODNet acceptance for AAAI 2022 !!! ( I firts read your MODNet paper about 1 year ago, and I...

Hey, great work on MODNet! I'm looking to try out modnet with a HRNet backbone for better results. Would you happen to already have the model function for HRNet that...

大佬,请问一下,官网体验效果比开源的模型推理效果要好很多,官网模型和开源模型是同一个网络结构,只是不同模型吗@ @ZHKKKe

作者你好 : 我发现有没有使用 pre-trained model 训练出来的效果差很大, 您之前说是使用了 https://github.com/thuyngch/Human-Segmentation-PyTorch 仓库中的预训练模型。在加载预训练模型后直接进行端到端训练 对此我有几个问题 1. 是直接使用他仓库上的预训练模型吗?还是使用他的网路另外训练你的数据的预训练模型? 2. 在 thuyngch/Human-Segmentation-PyTorch 仓库中没有看到 mobilenetv2_human_seg.ckpt 文件, 请问您仓库中所提供的 mobilenetv2_human_seg.ckpt 文件是如何产生的? 谢谢