liuheng92/tensorflow_PSENet

Issues

• No module named 'pse.pse

  #93 opened a year ago by vidkr
  0

• Tensorflow 2.0 not supported

  #92 opened 3 years ago by humbleOldSage
  0

• 请问为什么训练加载图片后一直是这样的

  #91 opened 4 years ago by whgao97
  0

• The author's code mystifies me！

  #82 opened 4 years ago by whereitogo
  1

• 修改data_provider.py来支持多边形格式输入--报错

  #75 opened 5 years ago by wuzuowuyou
  2

• Training loss not changing

  #87 opened 4 years ago by uraibeef
  3

• tensorflow_PSENet-master/pse/pse.so: undefined symbol: _Py_ZeroStruct

  #48 opened 5 years ago by banbsyip
  2

• 哪位有测试带recall和precision的脚本？

  #90 opened 4 years ago by SkiffDong
  1

• How to calculate recall or f-score?

  #89 opened 4 years ago by shwaitkumar
  1

• tf.train.get_checkpoint_state('/home/yons/models') 加载模型不起作用？？

  #88 opened 4 years ago by caozhongbjtu
  1

• 模型训练无法得到README中的结果

  #83 opened 4 years ago by zhangshaoyong
  1

• 我建了个QQ群，1080729300，大家一起来交流啊

  #81 opened 4 years ago by ingale726
  0

• 标签中的True和False不太理解

  #85 opened 4 years ago by Ri-Yue-Shui-Hua
  1

• 请问损失函数Lc中M，你是否使用OHEM，体现在哪里？

  #86 opened 4 years ago by firstadam
  1

• python3如何编译pse模块

  #84 opened 4 years ago by lx-rookie
  0

• 编译pse.so报错

  #80 opened 4 years ago by andy7166
  5

• 使用pretrained模型在icdar2015上复现出来结果不一致

  #58 opened 5 years ago by alexwangmac
  5

• 训练卡死，GPU利用率为0，但是显存占满

  #60 opened 5 years ago by cyLVlyl
  3

• model_loss异常

  #79 opened 4 years ago by yanqi1811
  0

• Can this model be trained and tested on Windows without GPU？

  #78 opened 5 years ago by zhong110020
  0

• AttributeError: 'NoneType' object has no attribute 'model_checkpoint_path'

  #77 opened 5 years ago by shwaitkumar
  1

• 训练不收敛，model loss 一直在0.3和0.8之间震荡

  #76 opened 5 years ago by asonee
  0

• Segmentation fault: 11

  #74 opened 5 years ago by mrwu-mac
  0

• Fatal Python error: PyThreadState_Get: no current thread

  #59 opened 5 years ago by RobertLeeL
  0

• 读取txt文本时，报编码格式的问题，

  #66 opened 5 years ago by guoshu6699
  1

• psenet标注用什么标注工具比较合适

  #70 opened 5 years ago by JinMana
  1

• 服务器显存不够，有哪些办法可以减小网络的大小阿，ResNet_v150可以改成小一点的VGG16么？

  #69 opened 5 years ago by apple2333cream
  1

• 同尺寸图像生成seg_maps耗时有10倍之差？

  #72 opened 5 years ago by pingzi5233
  0

• Deep deterministic policy gradient actor-critic algorithm with supervision controller

  #73 opened 5 years ago by Emercy44
  0

• 关于预训练模型

  #68 opened 5 years ago by zq2323
  1

• 请问有人尝试过利用自己数据进行训练时一张图片输入多个检测框的么？

  #67 opened 5 years ago by pingzi5233
  1

• CPU Support

  #65 opened 5 years ago by csmcallister
  1

• 训练问题

  #63 opened 5 years ago by amyjim9907
  3

• 输出[6,7,640,640]

  #64 opened 5 years ago by jun214384468
  2

• 请问博主训练了多少步呢？

  #62 opened 5 years ago by cyLVlyl
  1

• 数据读取问题

  #57 opened 5 years ago by ssvicnent
  1

• dice coefficient与原论文不一致

  #61 opened 5 years ago by ssvicnent
  1

• 单GPU训练PSENet，CPU占用将近100%，GPU显存占用不到2G，训练4天了一直无任何进展，也不报错

  #56 opened 5 years ago by SimonWang00
  2

• How to load pre_trained model?

  #55 opened 5 years ago by ceciliaAI
  2

• 训练复现问题、

  #54 opened 5 years ago by zhangxiaohuixuhao
  0

• How to get text out of boudry box

  #50 opened 5 years ago by nbhupendra
  1

• 请问eval生成的文本文件是什么，为什么会有浮点数，应该如何从文本提取出检测到的图片块呢~

  #47 opened 5 years ago by Leedorado
  1

• Error while running script please help!

  #49 opened 5 years ago by nbhupendra
  2

• when can I stop the training

  #52 opened 5 years ago by li10141110
  2

• 请问怎么用自己的数据集进行训练？

  #51 opened 5 years ago by GaffeyFF
  1

• 'NoneType' object has no attribute 'model_checkpoint_path'

  #46 opened 5 years ago by banbsyip
  10

• The variable is_edge does not change in previous loop in file pse.cpp

  #53 opened 5 years ago by minizon
  1

• 请问model loss和total loss分别降到多少可以停止训练，另外，训练步数一定要100000步吗，感觉得好久才能跑完

  #44 opened 5 years ago by Leedorado
  3

• FileNotFoundError: [WinError 2] 系统找不到指定的文件。

  #45 opened 5 years ago by banbsyip
  1

• 请问data_provider中的crop_area作用是什么？是否可以省去这一步？thx

  #43 opened 5 years ago by Leedorado
  0