[train hyper-parameters: Namespace(batch_size=4, epochs=40, lr=0.005, lrf=0.01)]
[epoch: 1]
global correct: 0.9836
precision: ['0.9836', '0.0000', '1.0000', '0.0000']
recall: ['1.0000', '0.0000', '0.0000', '0.0000']
IoU: ['0.9836', '0.0000', '0.0000', '0.0000']
mean IoU: 0.2459
[epoch: 2]
global correct: 0.9846
precision: ['0.9877', '0.0000', '0.4567', '0.3667']
recall: ['0.9982', '0.0000', '0.3550', '0.0001']
IoU: ['0.9860', '0.0000', '0.2496', '0.0001']
mean IoU: 0.3089
[epoch: 3]
global correct: 0.9759
precision: ['0.9959', '0.0000', '0.2372', '0.3812']
recall: ['0.9856', '0.0000', '0.7791', '0.0542']
IoU: ['0.9817', '0.0000', '0.2222', '0.0498']
mean IoU: 0.3134
[epoch: 4]
global correct: 0.9839
precision: ['0.9912', '0.0000', '0.4079', '0.5714']
recall: ['0.9951', '0.0000', '0.6690', '0.0002']
IoU: ['0.9864', '0.0000', '0.3394', '0.0002']
mean IoU: 0.3315
[epoch: 5]
global correct: 0.9841
precision: ['0.9953', '0.0000', '0.5245', '0.4968']
recall: ['0.9886', '0.0000', '0.7902', '0.6507']
IoU: ['0.9840', '0.0000', '0.4604', '0.3922']
mean IoU: 0.4591
[epoch: 6]
global correct: 0.9840
precision: ['0.9960', '0.0000', '0.5862', '0.4587']
recall: ['0.9878', '0.0000', '0.7291', '0.7869']
IoU: ['0.9839', '0.0000', '0.4813', '0.4080']
mean IoU: 0.4683
[epoch: 7]
global correct: 0.9806
precision: ['0.9976', '0.0000', '0.4699', '0.4352']
recall: ['0.9828', '0.0000', '0.8672', '0.8378']
IoU: ['0.9804', '0.0000', '0.4384', '0.4014']
mean IoU: 0.4551
[epoch: 8]
global correct: 0.9862
precision: ['0.9950', '0.0000', '0.5467', '0.5853']
recall: ['0.9910', '0.0000', '0.8076', '0.6035']
IoU: ['0.9861', '0.0000', '0.4837', '0.4227']
mean IoU: 0.4731
[epoch: 9]
global correct: 0.9852
precision: ['0.9968', '0.0000', '0.5522', '0.5141']
recall: ['0.9882', '0.0000', '0.8176', '0.7956']
IoU: ['0.9851', '0.0000', '0.4917', '0.4541']
mean IoU: 0.4827
[epoch: 10]
global correct: 0.9867
precision: ['0.9960', '0.0000', '0.5505', '0.5929']
recall: ['0.9905', '0.0000', '0.8339', '0.6890']
IoU: ['0.9866', '0.0000', '0.4961', '0.4677']
mean IoU: 0.4876
[epoch: 11]
global correct: 0.9872
precision: ['0.9959', '0.0000', '0.5838', '0.5876']
recall: ['0.9911', '0.0000', '0.7808', '0.7282']
IoU: ['0.9871', '0.0000', '0.5016', '0.4819']
mean IoU: 0.4926
[epoch: 12]
global correct: 0.9831
precision: ['0.9977', '0.0000', '0.4977', '0.4851']
recall: ['0.9852', '0.0000', '0.8838', '0.8397']
IoU: ['0.9830', '0.0000', '0.4672', '0.4439']
mean IoU: 0.4735
[epoch: 13]
global correct: 0.9866
precision: ['0.9967', '0.0000', '0.5614', '0.5651']
recall: ['0.9898', '0.0000', '0.8386', '0.7595']
IoU: ['0.9865', '0.0000', '0.5067', '0.4793']
mean IoU: 0.4931
[epoch: 14]
global correct: 0.9844
precision: ['0.9976', '0.0000', '0.5432', '0.4902']
recall: ['0.9865', '0.0000', '0.8625', '0.8502']
IoU: ['0.9842', '0.0000', '0.4999', '0.4512']
mean IoU: 0.4838
[epoch: 15]
global correct: 0.9818
precision: ['0.9984', '0.0000', '0.4797', '0.4619']
recall: ['0.9832', '0.0000', '0.9144', '0.8844']
IoU: ['0.9816', '0.0000', '0.4591', '0.4356']
mean IoU: 0.4691
[epoch: 16]
global correct: 0.9840
precision: ['0.9978', '0.0000', '0.5365', '0.4824']
recall: ['0.9860', '0.0000', '0.8717', '0.8614']
IoU: ['0.9838', '0.0000', '0.4972', '0.4477']
mean IoU: 0.4822
[epoch: 17]
global correct: 0.9869
precision: ['0.9968', '0.0000', '0.5902', '0.5556']
recall: ['0.9899', '0.0000', '0.8266', '0.7914']
IoU: ['0.9868', '0.0000', '0.5252', '0.4846']
mean IoU: 0.4992
[epoch: 18]
global correct: 0.9837
precision: ['0.9980', '0.0000', '0.5213', '0.4844']
recall: ['0.9855', '0.0000', '0.8898', '0.8638']
IoU: ['0.9836', '0.0000', '0.4897', '0.4500']
mean IoU: 0.4808
[epoch: 19]
global correct: 0.9825
precision: ['0.9982', '0.0000', '0.4990', '0.4663']
recall: ['0.9840', '0.0000', '0.9057', '0.8772']
IoU: ['0.9823', '0.0000', '0.4744', '0.4377']
mean IoU: 0.4736
[epoch: 20]
global correct: 0.9857
precision: ['0.9975', '0.0000', '0.5674', '0.5170']
recall: ['0.9880', '0.0000', '0.8547', '0.8394']
IoU: ['0.9855', '0.0000', '0.5175', '0.4705']
mean IoU: 0.4934
[epoch: 21]
global correct: 0.9863
precision: ['0.9973', '0.0000', '0.5858', '0.5293']
recall: ['0.9889', '0.0000', '0.8381', '0.8311']
IoU: ['0.9862', '0.0000', '0.5263', '0.4779']
mean IoU: 0.4976
[epoch: 22]
global correct: 0.9853
precision: ['0.9975', '0.0000', '0.5425', '0.5216']
recall: ['0.9876', '0.0000', '0.8631', '0.8291']
IoU: ['0.9851', '0.0000', '0.4996', '0.4710']
mean IoU: 0.4889
[epoch: 23]
global correct: 0.9856
precision: ['0.9974', '0.0000', '0.5268', '0.5495']
recall: ['0.9881', '0.0000', '0.8863', '0.7930']
IoU: ['0.9855', '0.0000', '0.4935', '0.4806']
mean IoU: 0.4899
[epoch: 24]
global correct: 0.9847
precision: ['0.9978', '0.0000', '0.5305', '0.5088']
recall: ['0.9867', '0.0000', '0.8875', '0.8437']
IoU: ['0.9845', '0.0000', '0.4971', '0.4650']
mean IoU: 0.4866
[epoch: 25]
global correct: 0.9848
precision: ['0.9978', '0.0000', '0.5369', '0.5081']
recall: ['0.9868', '0.0000', '0.8819', '0.8466']
IoU: ['0.9847', '0.0000', '0.5009', '0.4652']
mean IoU: 0.4877
[epoch: 26]
global correct: 0.9853
precision: ['0.9975', '0.0000', '0.5247', '0.5399']
recall: ['0.9876', '0.0000', '0.8884', '0.8126']
IoU: ['0.9852', '0.0000', '0.4922', '0.4801']
mean IoU: 0.4894
[epoch: 27]
global correct: 0.9866
precision: ['0.9971', '0.0000', '0.5662', '0.5575']
recall: ['0.9894', '0.0000', '0.8535', '0.7933']
IoU: ['0.9865', '0.0000', '0.5160', '0.4868']
mean IoU: 0.4973
[epoch: 28]
global correct: 0.9870
precision: ['0.9968', '0.0000', '0.6004', '0.5507']
recall: ['0.9901', '0.0000', '0.8019', '0.8047']
IoU: ['0.9869', '0.0000', '0.5229', '0.4858']
mean IoU: 0.4989
[epoch: 29]
global correct: 0.9855
precision: ['0.9975', '0.0000', '0.5421', '0.5312']
recall: ['0.9878', '0.0000', '0.8728', '0.8276']
IoU: ['0.9854', '0.0000', '0.5024', '0.4783']
mean IoU: 0.4915
[epoch: 30]
global correct: 0.9866
precision: ['0.9972', '0.0000', '0.5707', '0.5521']
recall: ['0.9892', '0.0000', '0.8535', '0.8074']
IoU: ['0.9865', '0.0000', '0.5198', '0.4879']
mean IoU: 0.4985
[epoch: 31]
global correct: 0.9853
precision: ['0.9975', '0.0000', '0.5133', '0.5519']
recall: ['0.9876', '0.0000', '0.9003', '0.8013']
IoU: ['0.9852', '0.0000', '0.4857', '0.4855']
mean IoU: 0.4891
[epoch: 32]
global correct: 0.9868
precision: ['0.9970', '0.0000', '0.5838', '0.5541']
recall: ['0.9897', '0.0000', '0.8375', '0.7956']
IoU: ['0.9867', '0.0000', '0.5244', '0.4850']
mean IoU: 0.4990
[epoch: 33]
global correct: 0.9869
precision: ['0.9969', '0.0000', '0.5747', '0.5671']
recall: ['0.9898', '0.0000', '0.8483', '0.7833']
IoU: ['0.9868', '0.0000', '0.5211', '0.4902']
mean IoU: 0.4995
[epoch: 34]
global correct: 0.9867
precision: ['0.9970', '0.0000', '0.5616', '0.5692']
recall: ['0.9896', '0.0000', '0.8622', '0.7778']
IoU: ['0.9866', '0.0000', '0.5154', '0.4896']
mean IoU: 0.4979
[epoch: 35]
global correct: 0.9868
precision: ['0.9970', '0.0000', '0.5763', '0.5566']
recall: ['0.9896', '0.0000', '0.8482', '0.7942']
IoU: ['0.9867', '0.0000', '0.5224', '0.4864']
mean IoU: 0.4989
[epoch: 36]
global correct: 0.9868
precision: ['0.9970', '0.0000', '0.5772', '0.5593']
recall: ['0.9896', '0.0000', '0.8484', '0.7940']
IoU: ['0.9867', '0.0000', '0.5232', '0.4884']
mean IoU: 0.4996
[epoch: 37]
global correct: 0.9858
precision: ['0.9976', '0.0000', '0.5537', '0.5310']
recall: ['0.9880', '0.0000', '0.8710', '0.8331']
IoU: ['0.9857', '0.0000', '0.5117', '0.4799']
mean IoU: 0.4943
[epoch: 38]
global correct: 0.9863
precision: ['0.9973', '0.0000', '0.5588', '0.5496']
recall: ['0.9888', '0.0000', '0.8665', '0.8101']
IoU: ['0.9862', '0.0000', '0.5145', '0.4869']
mean IoU: 0.4969
[epoch: 39]
global correct: 0.9864
precision: ['0.9972', '0.0000', '0.5644', '0.5505']
recall: ['0.9890', '0.0000', '0.8600', '0.8080']
IoU: ['0.9863', '0.0000', '0.5169', '0.4868']
mean IoU: 0.4975
[epoch: 40]
global correct: 0.9867
precision: ['0.9971', '0.0000', '0.5710', '0.5562']
recall: ['0.9894', '0.0000', '0.8531', '0.
没有合适的资源?快使用搜索试试~ 我知道了~
基于Swin-Transformer和Unet 项目、自适应多尺度训练、多类别分割:裂缝分割实战
共2000个文件
png:1692个
jpg:291个
py:8个
1.该资源内容由用户上传,如若侵权请联系客服进行举报
2.虚拟产品一经售出概不退款(资源遇到问题,请及时私信上传者)
2.虚拟产品一经售出概不退款(资源遇到问题,请及时私信上传者)
版权申诉
0 下载量 118 浏览量
2024-09-04
13:44:21
上传
评论
收藏 260.74MB 7Z 举报
温馨提示
基于Swin-Transformer和Unet 项目、自适应多尺度训练、多类别分割:裂缝分割实战 项目介绍:总大小260MB 本项目数据集:裂缝2类别分割数据集 网络仅仅训练了40个epochs,全局像素点的准确度达到0.98,训练epoch加大的话,性能还会更加优越! 代码介绍: 【训练】train 脚本会自动训练,代码会自动将数据随机缩放为设定尺寸的0.5-1.5倍之间,实现多尺度训练。为了实现多分割项目,utils中的compute_gray函数会将mask灰度值保存在txt文本,并且自动为网络定义输出的channel 【介绍】学习率采用cos衰减,训练集和测试集的损失和iou曲线可以在run_results文件内查看,图像由matplotlib库绘制。除此外,还保存了训练日志,最好权重等,在训练日志可以看到每个类别的iou、recall、precision以及全局像素点的准确率等等 【推理】把待推理图像放在inference目录下,直接运行predict脚本即可,无需设定参数 具体参考README文件,小白均可使用
资源推荐
资源详情
资源评论
收起资源包目录
基于Swin-Transformer和Unet 项目、自适应多尺度训练、多类别分割:裂缝分割实战 (2000个子文件)
1633_jpg.rf.b9028badb684eef34c91a4b4b9e48810.jpg 84KB
1633_jpg.rf.6fd036c3e99cd847acbd24ac64285477.jpg 83KB
1633_jpg.rf.e5d6848472b3dfdfc04690e1246b1f09.jpg 82KB
1607_jpg.rf.e024aac029f2ff399b3014046ab0c428.jpg 79KB
141_jpg.rf.dd4c8bdb012f04f4b1b2aed5db5566ed.jpg 79KB
1814_jpg.rf.3212be928968d070b8edb8649a2c4130.jpg 77KB
1783_jpg.rf.e890f64e2ea64721ac44241d4bbfbffe.jpg 77KB
1824_jpg.rf.643e1f0774e9b338d0735a001fbb6eb9.jpg 76KB
141_jpg.rf.7716261282faa71c0fcfbd9e7c22992a.jpg 76KB
1814_jpg.rf.e10fb25fc7de5e0f1f89c07f401824e2.jpg 76KB
1633_jpg.rf.cce2c2a824530e62d318e224aef60f21.jpg 75KB
1783_jpg.rf.08c4eccbc2fa08f60063b52173d20c75.jpg 74KB
1802_jpg.rf.f75c6e7286187907d232ea166af7560b.jpg 74KB
1876_jpg.rf.767772a310959e064af18b77e04614a4.jpg 74KB
141_jpg.rf.b124d3761d67c4b90adbd17a2d68657a.jpg 74KB
1814_jpg.rf.35121dc1689bd66a7d647b105d7471c5.jpg 73KB
1814_jpg.rf.9875cb37a9ef5f9a151bbfa3b0d36a85.jpg 73KB
1796_jpg.rf.5ca07a9a97f630ecabc8aecaf8c7e247.jpg 73KB
1802_jpg.rf.61c1a7bf0aa24d0750c1f6c013dabc9c.jpg 72KB
1803_jpg.rf.66aba4fe033375326cd13a2f03cc5ba6.jpg 72KB
1803_jpg.rf.cb1ff7feabe1f2bd4da513a5804598cb.jpg 71KB
1803_jpg.rf.3279b402c673e19cb02b5f6cc7ea2cbc.jpg 71KB
b.jpg 71KB
141_jpg.rf.7e2a7dc5202c60355f733cb15b6022fb.jpg 71KB
1796_jpg.rf.d655899039edf7e8c77308b8b5d887a9.jpg 70KB
1782_jpg.rf.f9c34fefbe3e4865e60d28eb14c7fe13.jpg 69KB
1876_jpg.rf.9fa42656aeb23d527ad382ce04161237.jpg 68KB
1782_jpg.rf.199f223eac4a05bc08f6dcaec0b5cb29.jpg 68KB
1669_jpg.rf.e128ec3a507a0f71368b5e20ad302bc0.jpg 68KB
1701_jpg.rf.8ba04f3e64c08fdf56cd5fc0aa929dc9.jpg 68KB
1833_jpg.rf.74bebc1393e53ac8b105d9231968b489.jpg 67KB
1690_jpg.rf.42dd8bd531e9d9f752c80f004b9f1678.jpg 67KB
1212_jpg.rf.e50e88174f24b4fb7591096aeb719513.jpg 67KB
1212_jpg.rf.6c5b116b5d50d8ff77e58651be47182e.jpg 67KB
1650_jpg.rf.93f1ff9ab7b1437b02fd82f27d92508d.jpg 66KB
1650_jpg.rf.b3bb881bc446ed2cbe3a224a34c4d16a.jpg 65KB
1613_jpg.rf.467cc6a69d6be750eb98e20598894544.jpg 64KB
1833_jpg.rf.1420343f0c05b83ece3e88cce89fb9a6.jpg 63KB
1807_jpg.rf.5ae5110318d855ba51b041d2703af30a.jpg 63KB
1864_jpg.rf.1d4fff1c943e488e33788eb8b9c4a9d8.jpg 63KB
1212_jpg.rf.bda22f52277fae01257bbc967ee781f1.jpg 62KB
1829_jpg.rf.28baaeb2f9d21ed4ec167dee08a800e0.jpg 62KB
1613_jpg.rf.450396c13138464caeba8206c4f80766.jpg 62KB
1803_jpg.rf.edd3e5135228c967060d281bf39b1c62.jpg 62KB
1807_jpg.rf.77201e616276f0b10fc9e2cab26a6a3c.jpg 62KB
1212_jpg.rf.15e3835d586dade38dc88eebc7e2afa4.jpg 62KB
1613_jpg.rf.e3bbda6f11c43c7fec7646da84c3cc52.jpg 61KB
1857_jpg.rf.65f81c65952d572811884516937213a3.jpg 61KB
1868_jpg.rf.db307be0f414891d55c09e84365687c4.jpg 60KB
1868_jpg.rf.2f73745838899ddecab2626a673c9f2e.jpg 60KB
1700_jpg.rf.7c4c2ab2e16661bace807bbd581b2f32.jpg 60KB
1853_jpg.rf.af8af71eea17ab42282a8207168f64c3.jpg 59KB
1850_jpg.rf.62cf625d166247f970824c78a159e687.jpg 59KB
1807_jpg.rf.02283bdcfdd04a4327092c27cbc37343.jpg 59KB
1614_jpg.rf.b262e1755b988985b57d43c387474143.jpg 59KB
1700_jpg.rf.ccff93a25fe88e8a728e8d3d0cc81018.jpg 59KB
1850_jpg.rf.7976f061272a50888c4b288e85532ba5.jpg 59KB
1676_jpg.rf.b7ce5202d783b278e844b6c289a37e68.jpg 59KB
1868_jpg.rf.8a4226bc383aea1731a5a38f9fe511ca.jpg 59KB
1614_jpg.rf.23617cf93ab294ef3b698f21ea7f4257.jpg 58KB
1212_jpg.rf.3c2a47f4d86d03f603df0f324d57a690.jpg 58KB
1855_jpg.rf.398002df13fc4ddf44556a946d40443a.jpg 58KB
a.jpg 58KB
141_jpg.rf.0e83a190be816f3fc68c3cbeb7125baf.jpg 58KB
141_jpg.rf.edc32c2f0c148d4953c8be742bcf7233.jpg 58KB
1838_jpg.rf.8a5a7903e9bc634a00410d6d8ed8754f.jpg 58KB
1827_jpg.rf.e296fabf0a19ce364912b23b22382376.jpg 58KB
1827_jpg.rf.f4044cf26e01cb6c97cd1e8f64d4e987.jpg 57KB
1827_jpg.rf.61c7d7379b3d68888d19caeac3c33bf5.jpg 57KB
1793_jpg.rf.000a31e0ae3a865998f31d3f86199173.jpg 56KB
1231_jpg.rf.4ff957fb8bf5641a4ee1baee683714e0.jpg 56KB
1800_jpg.rf.42f7fba142d516dafc02db213a48fe20.jpg 56KB
1793_jpg.rf.45916142ecec7f5a8acf66014167a154.jpg 56KB
1800_jpg.rf.99622087cb0a3be64a6860a52f0b91f7.jpg 56KB
1793_jpg.rf.27da06e7f26be4697457e4174ce4a88b.jpg 56KB
1877_jpg.rf.fc91736ddc8ad37f61a9218dcbbc48d2.jpg 56KB
1789_jpg.rf.bd8580af7b7dc49392b0e0894d5efa47.jpg 56KB
1859_jpg.rf.ce7744e6eb218fb84154d06ec0db55d3.jpg 55KB
1789_jpg.rf.c6a04271abe98cd7d7e4b37af811bc2c.jpg 55KB
1827_jpg.rf.887cb6b88940e3fd2133f44ff9c3fa15.jpg 55KB
1789_jpg.rf.c726636b5f002ee56cc2110c2bc40113.jpg 55KB
1859_jpg.rf.e007f9dc45dff6e3dda0d33421140f31.jpg 55KB
1700_jpg.rf.518f860243e3bf0775876bedc4f2e9d1.jpg 55KB
1793_jpg.rf.c6f0efc6aea6005796872240a919ac12.jpg 55KB
1859_jpg.rf.4548c370f60cfd6fe7476c101262ed51.jpg 55KB
1877_jpg.rf.21d2259af05627bb28101e25bf345117.jpg 55KB
1829_jpg.rf.e0b73ba638342307a7a69617baf354b7.jpg 55KB
1877_jpg.rf.c88e3b840a97ef1c6708071019aaebfe.jpg 55KB
1877_jpg.rf.386b94b9ac92a05e5be58267fb625339.jpg 54KB
1827_jpg.rf.a3b3532d47a612b79b7bdad188e91f8f.jpg 54KB
1862_jpg.rf.77735e07512d65787ee93a0791d8566a.jpg 54KB
1630_jpg.rf.e71b7b32d481b9033f30e8b232d74a90.jpg 54KB
1827_jpg.rf.8fbb35ee6572cf21a7d0c90b8b9ba680.jpg 54KB
1789_jpg.rf.8540d4e7c2b0db9b9f33d1ce2b284332.jpg 53KB
1664_jpg.rf.b2a5d756399228f2495a241bd1f44854.jpg 53KB
1800_jpg.rf.8b0fc30471e8a7d1c88be47a6a519ba8.jpg 53KB
1664_jpg.rf.feddbf3e7ac7b4bbcb2cd152ca193b77.jpg 53KB
1664_jpg.rf.079f37e70430f17143b0c8c6e5ce036e.jpg 53KB
1676_jpg.rf.fb081de2279c6e318520515a15348b09.jpg 53KB
1664_jpg.rf.ba4121e43eee93e9ecb99f2ce181bc4e.jpg 53KB
共 2000 条
- 1
- 2
- 3
- 4
- 5
- 6
- 20
资源评论
Ai医学图像分割
- 粉丝: 2w+
- 资源: 2128
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
最新资源
- vmware-VMnet8一键启动和停止脚本
- 可移植的 Python 数据框库.zip
- 包含 Andrei Neagoie 的《从零到精通掌握编码面试 - 数据结构 + 算法》课程的所有代码示例,使用 Python 语言 .zip
- 数据库课程设计(图书馆管理系统)springboot+swing+mysql+mybatis
- C++ Vigenère 密码(解密代码)
- zblog日收站群,zblog泛目录
- C++ Vigenère 密码(加密代码)
- Vue Router 是 Vue 生态系统的一部分,是一个 MIT 许可的开源项目,其持续开发完全在赞助商的支持下成为可能 支持 Vue 路由器
- PM2.5 数据集 包含上海、成都、广州、北京、沈阳五地的PM2.5观测,csv文件
- 电动汽车与软件定义汽车(SDV)时代的汽车行业数字化转型
资源上传下载、课程学习等过程中有任何疑问或建议,欢迎提出宝贵意见哦~我们会及时处理!
点击此处反馈
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功