xuebinqin/U-2-Net issues and pull requests

#391 - How to remove the inner part of the product?

Issue - State: open - Opened by html5maker 2 months ago - 4 comments

#390 - Removes the part of the product instead of background

Issue - State: open - Opened by html5maker 2 months ago

#389 - Request for Permission to Use Salient Object Detection Prediction Images and Digital Signature

Issue - State: open - Opened by peenwin 3 months ago - 1 comment

#388 - remove background with U-2-Net is not as expected,what should I do

Issue - State: open - Opened by skyyan2000 4 months ago

#387 - 如何实现多种分类

Issue - State: open - Opened by hhhzr 5 months ago - 1 comment

#386 - Feature/white background

Pull Request - State: open - Opened by qmfw 5 months ago

#385 - How can i use my own GPU to train

Issue - State: open - Opened by Dan-H-Muniz-Sanchez 6 months ago

#384 - Paddlehub is no longer under maintenance, what should I do if an error message is installed

Issue - State: open - Opened by itmeicn 6 months ago

#383 - 自己数据训练的u2net，归一化尺寸320，但是抠图效果边缘锯齿明显

Issue - State: closed - Opened by xiangfeidecao 6 months ago

#382 - u2net for semantic segmentation

Issue - State: open - Opened by fhfeishi 7 months ago

#381 - num_workers=1

Issue - State: open - Opened by wangfugui127 8 months ago

#380 - 有什么办法只保留抠图部分尺寸，不需要原图尺寸大小，谢谢

Issue - State: open - Opened by zhoujian1028 9 months ago

#379 - The results are not good when the object has small holes.

Issue - State: open - Opened by Wananrd 10 months ago

#378 - can tell me , label mask must be 0 or 1,if 0 or 1，why binary is smooth, if dataset mask edge in (0,1),it can work?

Issue - State: open - Opened by Trway 10 months ago

#377 - the effect of replacing background with U-2-Net is not as expected, anyone can offer a help

Issue - State: open - Opened by weiweiwang 10 months ago

#376 - u2net_train.py returning ValueError

Issue - State: closed - Opened by rajmanna-dev 10 months ago - 1 comment

#375 - /alueror: At least one stride in the given numpy aray is mepative, ad tensors with nepative strides are not curently suported. (You can probably work around this by making a copy of your array with array.copy( ).)

Issue - State: open - Opened by liuyisu 10 months ago

#374 - 请问大家，我需要前景而非mask，该怎样修改代码呢，感谢

Issue - State: closed - Opened by liuyisu 11 months ago

#373 - Usage on cross platform mobile devices using tenser flow or pytourch ( Special on Flutter )

Issue - State: open - Opened by sadaqatdev 11 months ago

#372 - Can' Access Human Segmentation Model Weights

Issue - State: open - Opened by thisisashukla 11 months ago - 1 comment

#371 - ImportError: cannot import name 'U2NET' from 'model'

Issue - State: open - Opened by tzktz 12 months ago - 2 comments

#370 - 如何添加评价指标

Issue - State: closed - Opened by MayLuoA 12 months ago

#369 - fix code

Pull Request - State: open - Opened by tdp1996 almost 1 year ago

#368 - Remove pycache & Add .gitignore

Pull Request - State: open - Opened by pooya-mohammadi about 1 year ago

#367 - Is it support person segmentation now?

Issue - State: open - Opened by magic3584 about 1 year ago - 2 comments

#366 - Bump pillow from 8.1.1 to 10.0.1

Pull Request - State: open - Opened by dependabot[bot] about 1 year ago
Labels: dependencies

#365 - 请问该模型只能输出二分类结果么？可以输出为多分类么？比如三分类

Issue - State: open - Opened by lhq521 about 1 year ago - 1 comment

#364 - 能否麻烦将整个项目打个包（包含运行文件和预训练模型），集成一个 .bat 文件，点击运行即可使用？

Issue - State: open - Opened by auendx about 1 year ago - 2 comments

#363 - How can I input video or webcam in the test.py script?

Issue - State: open - Opened by Hogushake about 1 year ago

#362 - 对u2netp模型进行qat量化

Issue - State: open - Opened by ZHIZIHUABU over 1 year ago

#361 - 模型训练时间过长

Issue - State: closed - Opened by ZHIZIHUABU over 1 year ago

#359 - Inference speed

Issue - State: open - Opened by cindymuji over 1 year ago - 1 comment

#358 - Continue training, train my own model with U-2-Net, in between due to some reasons the training was interrupted or I want to strengthen an existing model, what should I do? Can you provide a 'Continue-training.py'?

Issue - State: closed - Opened by panpengfei21 over 1 year ago - 4 comments

#357 - Hi, when I run this code, I get strange errors in other detection tasks.The following is the warning where the error occurs

Issue - State: open - Opened by 1dhuh over 1 year ago - 1 comment

#356 - RescaleT:Why not prioritize maintaining the aspect ratio

Issue - State: closed - Opened by zhouchengfei over 1 year ago

#355 - 大佬，可以提供一下验证模型相关指标的代码吗？

Issue - State: open - Opened by zhang-pan over 1 year ago - 1 comment

#354 - RuntimeError: unexpected EOF, The file might be corrupted

Issue - State: open - Opened by coloedrainbow over 1 year ago

#353 - Ai

Issue - State: open - Opened by saadben21 over 1 year ago

#352 - [Question] Trained model for proprietary use

Issue - State: open - Opened by Electr0phile almost 2 years ago - 1 comment

#351 - onnx result wrong

Issue - State: closed - Opened by NingNanXin almost 2 years ago - 5 comments

#350 - Cannot Import U2NET

Issue - State: open - Opened by FASTANDEXTREME almost 2 years ago

#349 - issue about portrait function

Issue - State: open - Opened by tomstevelfq almost 2 years ago

#348 - U-2-Net multi-gpu Training!

Issue - State: open - Opened by skuley almost 2 years ago - 2 comments

#347 - Add workflow

Pull Request - State: closed - Opened by davidwang-unity almost 2 years ago

#346 - Resucing latency

Issue - State: open - Opened by divyanshi00 almost 2 years ago - 1 comment

#345 - How to train the human portrait drawing model?

Issue - State: closed - Opened by Harryoung almost 2 years ago

#344 - data

Issue - State: open - Opened by qaf1344 almost 2 years ago

#343 - Inference model with MNN framework

Issue - State: closed - Opened by AnhPC03 almost 2 years ago - 1 comment

#342 - Calculating the Inference Metrics

Issue - State: open - Opened by vastech73 almost 2 years ago - 1 comment

#341 - Bump pillow from 8.1.1 to 9.3.0

Pull Request - State: closed - Opened by dependabot[bot] about 2 years ago - 1 comment
Labels: dependencies

#340 - How to get the evaluation metrics value?

Issue - State: open - Opened by AbrahamMulat about 2 years ago - 3 comments

#339 - TensorRT conversion produces different outputs

Issue - State: open - Opened by aviadmx about 2 years ago

#338 - Image to image translation

Issue - State: open - Opened by dongyun-kim-arch about 2 years ago

#337 - Improving results of foreground segmentation by using U2net

Issue - State: open - Opened by thebeautyofcoding about 2 years ago - 1 comment

#336 - A CUDA memory leak on models defined in u2net_refactor.py

Issue - State: open - Opened by chmendoza about 2 years ago

#335 - How to test using my trained model?

Issue - State: open - Opened by hic9507 about 2 years ago

#334 - How to plot accuray and loss graph?

Issue - State: open - Opened by hic9507 about 2 years ago - 1 comment

#333 - A new webapp that generate ID photos for free based on U2NET

Issue - State: open - Opened by hckuo about 2 years ago

#332 - Any advices on image restoration

Issue - State: open - Opened by anguoyang about 2 years ago

#331 - Should I flip the training dataset images or does your code do that?

Issue - State: open - Opened by alcinopitchprint about 2 years ago - 2 comments

#330 - Is it possible to create conditional version of model?

Issue - State: open - Opened by egorovivannn about 2 years ago

#329 - how get image matting

Issue - State: open - Opened by dh12306 about 2 years ago - 2 comments

#328 - How To Use RemoveBackground module

Issue - State: open - Opened by 1847427213 about 2 years ago

#327 - Image Segmentation using u2Net model android.

Issue - State: open - Opened by ahmadbajwa8282 about 2 years ago

#326 - The result isn't always accurate, how can I improve it ?

Issue - State: closed - Opened by X901 about 2 years ago - 1 comment

#325 - Training speed-up

Issue - State: open - Opened by dem-esgal about 2 years ago

#324 - a picture containing a closed figure

Issue - State: open - Opened by constantine008 over 2 years ago - 3 comments

#323 - no need Conv2d bias for Conv2d+BN module.

Issue - State: closed - Opened by WZMIAOMIAO over 2 years ago - 1 comment

#322 - Do you have a pre-trained model of u2netp for human portrait drawing？

Issue - State: open - Opened by zozoteacher over 2 years ago - 3 comments

#321 - What is the time complexity?

Issue - State: open - Opened by BennyTheDev over 2 years ago - 1 comment

#320 - The U-2-Net architecture predicts blurry "atypical" masks!

Issue - State: open - Opened by ashikrafi over 2 years ago - 2 comments

#319 - Hello, I'm a student of Tsinghua University. I found that running unet2_ test. The values of the image generated by py file are not all 0 and 255, but also some values between 0 and 255. I want to ask why. Usually, the segmentation result is not only two values, that is, 0 or 255

Issue - State: open - Opened by xianglei3 over 2 years ago - 2 comments

#318 - I am a student of the second year of Tsinghua graduate school. I sincerely ask a question: does it support real-time verification during training? For example, do you consider adding a new function, that is, each time you complete an epoch training, you will verify on the specified verification data set and print the loss on the verification set?

Issue - State: open - Opened by xianglei3 over 2 years ago - 2 comments

#317 - I'm a student of the second Graduate School of Tsinghua University. When I trained, I didn't quite understand the meaning of L1 to L5 loss printed out by the command line. Among them, do L1 and L2 loss refer to absolute value error and mean square error? In addition, I don't understand L3 to L5 loss, and I sincerely look forward to your reply

Issue - State: open - Opened by xianglei3 over 2 years ago - 1 comment

#316 - I am a student of Tsinghua University. I want to ask whether u2net only supports single category segmentation. For example, for the picture of people riding on bicycles, can people and bicycles be separated separately? Thank you. Thank you very much

Issue - State: open - Opened by xianglei3 over 2 years ago - 1 comment

#315 - 可以去背景吗，不止是人象

Issue - State: open - Opened by gg22mm over 2 years ago - 1 comment

#314 - Could you provide the saliency map on the SOD test datasets?such as ECSSD/DUT-TE/DUT-OMRON/PASCAL/HKU-IS? Thank you for your contribute to SOD field.

Issue - State: open - Opened by StormArcher over 2 years ago - 1 comment

#313 - Bump numpy from 1.15.2 to 1.22.0

Pull Request - State: open - Opened by dependabot[bot] over 2 years ago
Labels: dependencies

#312 - Mask蒙版图放大后的边界不清晰问题

Issue - State: open - Opened by jackiehu over 2 years ago - 1 comment

#311 - GPU memory increases during training

Issue - State: open - Opened by kenmbkr over 2 years ago - 1 comment

#310 - Please add Tags to find your Project

Issue - State: open - Opened by dazzafact over 2 years ago - 1 comment

#309 - BCELoss was unstable so just get logits from model and use BCEwithLog…

Pull Request - State: open - Opened by deshwalmahesh over 2 years ago

#308 - 3D Photo: New app on iOS using U-2-Net

Issue - State: open - Opened by adirkol over 2 years ago

#307 - reducing false positives with produce an all-black mask for images without annotations

Issue - State: open - Opened by mitrasafari over 2 years ago

#306 - Background removal module

Issue - State: open - Opened by kos94ok over 2 years ago - 2 comments

#305 - Training time is much higher

Issue - State: open - Opened by danishboy000 over 2 years ago - 2 comments

#304 - How to reduce occupied cuda when test images? [email protected]

Issue - State: open - Opened by nissansz over 2 years ago

#303 - how to draw such a beautiful network architecture picture?

Issue - State: closed - Opened by Breezewrf over 2 years ago - 2 comments

#302 - How to desaturate background. Help needed.

Issue - State: open - Opened by schwarzwals over 2 years ago - 4 comments

#301 - New Lensto App

Issue - State: closed - Opened by ioskevinshah over 2 years ago - 3 comments

#300 - how to adjust selection

Issue - State: open - Opened by schwarzwals over 2 years ago - 3 comments

#299 - how to see the accuracy from the training log? [email protected]

Issue - State: open - Opened by nissansz over 2 years ago

#298 - train error occurred. [email protected]

Issue - State: closed - Opened by nissansz over 2 years ago - 1 comment

#297 - Training from scratch

Issue - State: closed - Opened by denddyprod over 2 years ago - 1 comment

#296 - hi, how to train to process subtitle images? Any pretrained models? [email protected]. Want to get clearn white background, black characters in image

Issue - State: closed - Opened by nissansz over 2 years ago - 1 comment

#295 - Model size reduced to 43Mb! and new Webapp

Issue - State: open - Opened by Kikedao over 2 years ago - 32 comments

#294 - Bump pillow from 8.1.1 to 9.0.1

Pull Request - State: closed - Opened by dependabot[bot] over 2 years ago - 1 comment
Labels: dependencies

#293 - spatial size

Issue - State: open - Opened by ouyuxuanbridge over 2 years ago

#292 - label format

Issue - State: open - Opened by xiyufeng2 over 2 years ago - 1 comment

#291 - DUTS-HQ: Idea for U2Net(HQ)

Issue - State: open - Opened by dbpprt over 2 years ago

GitHub / xuebinqin/U-2-Net issues and pull requests