CCT:[CVPR 2020] 具有交叉一致性训练的半监督语义分割-源码

上传者: 42122306 | 上传时间: 2021-09-07 10:43:11 | 文件大小: 1.57MB | 文件类型: ZIP
具有交叉一致性训练 (CCT) 的半监督语义分割 , 本 repo 包含 CVPR 2020 论文的官方实现:Semi-Supervised Semantic Segmentation with Cross-Consistecy Training,它采用了传统的半监督学习的一致性训练框架进行语义分割,扩展到弱监督学习和在多个域。 强调 (1) 语义分割的一致性训练。 我们观察到,对于语义分割,由于任务的密集性,集群假设更容易在隐藏表示而不是输入上强制执行。 (2) 交叉培训。 我们为半监督语义分割提出了 CCT(Cross-Consistency Training),我们在其中定义了许多新的扰动,并展示了对编码器输出而不是输入执行一致性的有效性。 (3) 使用来自多个域的弱标签和像素级标签。 所提出的方法非常简单灵活,并且可以很容易地扩展到使用来自多个域的图像级标签和像素级标签。 要求

文件下载

资源详情

[{"title":"( 69 个子文件 1.57MB ) CCT:[CVPR 2020] 具有交叉一致性训练的半监督语义分割-源码","children":[{"title":"CCT-master","children":[{"title":"inference.py <span style='color:#111;'> 4.33KB </span>","children":null,"spread":false},{"title":"train.py <span style='color:#111;'> 3.16KB </span>","children":null,"spread":false},{"title":"utils","children":[{"title":"htmlwriter.py <span style='color:#111;'> 3.54KB </span>","children":null,"spread":false},{"title":"helpers.py <span style='color:#111;'> 1.94KB </span>","children":null,"spread":false},{"title":"ramps.py <span style='color:#111;'> 1.02KB </span>","children":null,"spread":false},{"title":"__init__.py <span style='color:#111;'> 26B </span>","children":null,"spread":false},{"title":"metrics.py <span style='color:#111;'> 3.27KB </span>","children":null,"spread":false},{"title":"losses.py <span style='color:#111;'> 11.48KB </span>","children":null,"spread":false},{"title":"lr_scheduler.py <span style='color:#111;'> 3.14KB </span>","children":null,"spread":false},{"title":"pallete.py <span style='color:#111;'> 541B </span>","children":null,"spread":false},{"title":"logger.py <span style='color:#111;'> 428B </span>","children":null,"spread":false}],"spread":true},{"title":"models","children":[{"title":"decoders.py <span style='color:#111;'> 8.87KB </span>","children":null,"spread":false},{"title":"model.py <span style='color:#111;'> 7.23KB </span>","children":null,"spread":false},{"title":"__init__.py <span style='color:#111;'> 22B </span>","children":null,"spread":false},{"title":"encoder.py <span style='color:#111;'> 2.42KB </span>","children":null,"spread":false},{"title":"backbones","children":[{"title":"get_pretrained_model.sh <span style='color:#111;'> 214B </span>","children":null,"spread":false},{"title":"__init__.py <span style='color:#111;'> 0B </span>","children":null,"spread":false},{"title":"resnet_models.py <span style='color:#111;'> 10.12KB </span>","children":null,"spread":false},{"title":"module_helper.py <span style='color:#111;'> 6.05KB </span>","children":null,"spread":false},{"title":"resnet_backbone.py <span style='color:#111;'> 6.52KB </span>","children":null,"spread":false}],"spread":true}],"spread":true},{"title":"configs","children":[{"title":"config.json <span style='color:#111;'> 2.35KB </span>","children":null,"spread":false}],"spread":true},{"title":"base","children":[{"title":"base_trainer.py <span style='color:#111;'> 7.16KB </span>","children":null,"spread":false},{"title":"__init__.py <span style='color:#111;'> 115B </span>","children":null,"spread":false},{"title":"base_dataloader.py <span style='color:#111;'> 1.57KB </span>","children":null,"spread":false},{"title":"base_dataset.py <span style='color:#111;'> 7.04KB </span>","children":null,"spread":false},{"title":"base_model.py <span style='color:#111;'> 862B </span>","children":null,"spread":false}],"spread":true},{"title":"requirements.txt <span style='color:#111;'> 208B </span>","children":null,"spread":false},{"title":"dataloaders","children":[{"title":"__init__.py <span style='color:#111;'> 20B </span>","children":null,"spread":false},{"title":"voc_splits","children":[{"title":"800_train_unsupervised.txt <span style='color:#111;'> 630.48KB </span>","children":null,"spread":false},{"title":"200_train_unsupervised.txt <span style='color:#111;'> 669.15KB </span>","children":null,"spread":false},{"title":"1464_train_supervised.txt <span style='color:#111;'> 94.36KB </span>","children":null,"spread":false},{"title":"1000_train_supervised.txt <span style='color:#111;'> 64.45KB </span>","children":null,"spread":false},{"title":"800_train_supervised.txt <span style='color:#111;'> 51.56KB </span>","children":null,"spread":false},{"title":"1000_train_unsupervised.txt <span style='color:#111;'> 617.59KB </span>","children":null,"spread":false},{"title":"500_train_supervised.txt <span style='color:#111;'> 32.23KB </span>","children":null,"spread":false},{"title":"boxes.json <span style='color:#111;'> 4.91MB </span>","children":null,"spread":false},{"title":"200_train_supervised.txt <span style='color:#111;'> 12.89KB </span>","children":null,"spread":false},{"title":"500_train_unsupervised.txt <span style='color:#111;'> 649.82KB </span>","children":null,"spread":false},{"title":"val.txt <span style='color:#111;'> 93.39KB </span>","children":null,"spread":false},{"title":"100_train_supervised.txt <span style='color:#111;'> 6.45KB </span>","children":null,"spread":false},{"title":"classes.json <span style='color:#111;'> 3.44MB </span>","children":null,"spread":false},{"title":"300_train_unsupervised.txt <span style='color:#111;'> 662.71KB </span>","children":null,"spread":false},{"title":"300_train_supervised.txt <span style='color:#111;'> 19.34KB </span>","children":null,"spread":false},{"title":"100_train_unsupervised.txt <span style='color:#111;'> 675.60KB </span>","children":null,"spread":false},{"title":"1464_train_unsupervised.txt <span style='color:#111;'> 587.68KB </span>","children":null,"spread":false},{"title":"60_train_supervised.txt <span style='color:#111;'> 3.87KB </span>","children":null,"spread":false},{"title":"60_train_unsupervised.txt <span style='color:#111;'> 678.18KB </span>","children":null,"spread":false}],"spread":false},{"title":"voc.py <span style='color:#111;'> 2.17KB </span>","children":null,"spread":false}],"spread":true},{"title":"LICENSE <span style='color:#111;'> 1.04KB </span>","children":null,"spread":false},{"title":"pseudo_labels","children":[{"title":"make_cam.py <span style='color:#111;'> 2.70KB </span>","children":null,"spread":false},{"title":"train_cam.py <span style='color:#111;'> 3.44KB </span>","children":null,"spread":false},{"title":"run.py <span style='color:#111;'> 2.42KB </span>","children":null,"spread":false},{"title":"voc12","children":[{"title":"dataloader.py <span style='color:#111;'> 9.07KB </span>","children":null,"spread":false},{"title":"test.txt <span style='color:#111;'> 17.06KB </span>","children":null,"spread":false},{"title":"train.txt <span style='color:#111;'> 17.16KB </span>","children":null,"spread":false},{"title":"cls_labels.npy <span style='color:#111;'> 2.32MB </span>","children":null,"spread":false},{"title":"val.txt <span style='color:#111;'> 16.98KB </span>","children":null,"spread":false},{"title":"make_cls_labels.py <span style='color:#111;'> 1007B </span>","children":null,"spread":false},{"title":"train_aug.txt <span style='color:#111;'> 124.01KB </span>","children":null,"spread":false}],"spread":false},{"title":"net","children":[{"title":"resnet50.py <span style='color:#111;'> 3.87KB </span>","children":null,"spread":false},{"title":"resnet50_cam.py <span style='color:#111;'> 1.73KB </span>","children":null,"spread":false}],"spread":false},{"title":"README.md <span style='color:#111;'> 1.80KB </span>","children":null,"spread":false},{"title":"cam_to_pseudo_labels.py <span style='color:#111;'> 2.04KB </span>","children":null,"spread":false},{"title":"misc","children":[{"title":"torchutils.py <span style='color:#111;'> 2.00KB </span>","children":null,"spread":false},{"title":"pyutils.py <span style='color:#111;'> 2.54KB </span>","children":null,"spread":false},{"title":"imutils.py <span style='color:#111;'> 8.07KB </span>","children":null,"spread":false}],"spread":false}],"spread":true},{"title":"README.md <span style='color:#111;'> 8.85KB </span>","children":null,"spread":false},{"title":".gitignore <span style='color:#111;'> 1.38KB </span>","children":null,"spread":false},{"title":"trainer.py <span style='color:#111;'> 12.50KB </span>","children":null,"spread":false}],"spread":false}],"spread":true}]

评论信息

免责申明

【只为小站】的资源来自网友分享,仅供学习研究,请务必在下载后24小时内给予删除,不得用于其他任何用途,否则后果自负。基于互联网的特殊性,【只为小站】 无法对用户传输的作品、信息、内容的权属或合法性、合规性、真实性、科学性、完整权、有效性等进行实质审查;无论 【只为小站】 经营者是否已进行审查,用户均应自行承担因其传输的作品、信息、内容而可能或已经产生的侵权或权属纠纷等法律责任。
本站所有资源不代表本站的观点或立场,基于网友分享,根据中国法律《信息网络传播权保护条例》第二十二条之规定,若资源存在侵权或相关问题请联系本站客服人员,zhiweidada#qq.com,请把#换成@,本站将给予最大的支持与配合,做到及时反馈和处理。关于更多版权及免责申明参见 版权及免责申明