bubbliiiing/mask-rcnn-keras

Issues

• 对于ROI Align层中的截取过程不太理解，希望博主能拨冗解答！！

  #54 opened 7 months ago by stariveryoung
  0

• 安装keras时出现依赖项冲突

  #53 opened 7 months ago by stariveryoung
  2

• 如何单独获得每个预测实例的分割结果

  #52 opened 9 months ago by svoeff
  0

• train.py:232: UserWarning: Output "ROI" missing from loss dictionary. We assume this was done on purpose, and we will not be expecting any data to be passed to "ROI" during training.

  #32 opened 3 years ago by CaoXiang-space
  4

• 请问可以在已有模型上继续训练么

  #7 opened 4 years ago by evilvlso
  5

• 应该与各自图中的真实框的类别号中验证pred_class_logits所指的实际类别号在各自图中的真实框的类别号中是否存在

  #51 opened a year ago
  0

• 你好！请问如果我把主干网改成resnet50，我应该去哪下载有关resnet50+FPN在COCO数据集上预训练的权重呢

  #50 opened a year ago by 00000022234
  1

• IndexError: boolean index did not match indexed array along dimension 0; dimension is 1 but corresponding boolean dimension is 8

  #49 opened a year ago by bankshotbandwidth
  1

• 预测出来的图片中并没有区分实例类别

  #40 opened 2 years ago by DarlingOhayo1
  3

• 不得不说效果很流批

  #48 opened 2 years ago by sparksWLL
  1

• coco数据集标签文件是自带的instances_val2017.json吗

  #47 opened 2 years ago by sparksWLL
  0

• 可以出一个pytorch版本的吗？

  #46 opened 2 years ago by pjw2146087
  0

• 测试的时候出现IndexError: list index out of range

  #45 opened 2 years ago by Pengfeicc
  2

• 复现失败

  #44 opened 2 years ago by 1996Eric
  4

• 调用多显卡时显示报错

  #43 opened 2 years ago by revquiet
  3

• ValueError: **slice index -1 of dimension 0 out of bounds.** for '{{node loss/rpn_class_loss_loss/strided_slice}} =

  #42 opened 2 years ago by immaping
  1

• 1000+图片,但仍显示数据集过小

  #41 opened 2 years ago by right135
  12

• tensorflow.python.framework.errors_impl.InvalidArgumentError: indices[1] = 130885 is not in [0, 65472)

  #39 opened 2 years ago by zDAIght
  24

• 开始训练 shape 数据集的错误

  #31 opened 3 years ago by xudaopao
  11

• 使用resnet50训练

  #38 opened 2 years ago by theoning
  2

• 训练时怎么只加载主干网络的权重？

  #37 opened 2 years ago by gfzwytc
  1

• 预测结果只有一类

  #36 opened 2 years ago by theoning
  1

• 预测结果很差

  #35 opened 2 years ago by apz27
  1

• tf2版本的代码

  #22 opened 3 years ago by codeseeking
  4

• 我导，在trian的时候，我设置的epoch为100，但是每次训练到epoch=3,4次的时候，就停止了，这是咋回事腻

  #33 opened 3 years ago by binghui10086
  7

• 我cuda版本太高了（11.4），导致tensorflow-gpu无法兼容

  #34 opened 3 years ago by 31901026sxc
  10

• 更换权重类型

  #30 opened 3 years ago by Ulricman
  1

• 训练coco数据集过程中，Exception: Invalid bounding box with area of zero，请问您知道怎么解决嘛

  #29 opened 3 years ago by KevinR10
  8

• 训练中的bug

  #28 opened 3 years ago by jiangzhenyudlut
  11

• 需要requirements.txt文件

  #26 opened 3 years ago by 20000131
  5

• num_workers在哪里修改

  #27 opened 3 years ago by 20000131
  2

• 请问之后会出pytorch版本的mask-rcnn吗

  #25 opened 3 years ago by Moris-Zhan
  1

• 从头训练

  #24 opened 3 years ago by Jasonx001
  1

• ValueError: None values not supported.

  #23 opened 3 years ago by super-ys
  2

• 我的cdua版本是10.0，tensorflow-gpu版本是1.13.1，但是cudnn没有找到与cunda10.0对应的cudnn7.4版本，有人遇到相同问题吗

  #13 opened 3 years ago by Jess1989
  1

• 图片中标注比较多，训练速度特别慢？请教下如何优化呢？

  #21 opened 3 years ago by jiangxinufo
  2

• 一张图有100多个实例 怎么提高速度训练啊 太慢了

  #6 opened 4 years ago by dlml
  5

• 求救bubbliiing大神：Cannot iterate over a shape with unknown rank.

  #20 opened 3 years ago by wt616
  1

• module 'tensorflow.math' has no attribute 'is_nan'

  #19 opened 3 years ago by fancyfeifei
  1

• 老大，请问下同张图片的不同角度能否准确识别

  #18 opened 3 years ago by yelynn
  2

• 有人试过转换成tflite吗？

  #17 opened 3 years ago by lbininhbl
  5

• ValueError: Cannot iterate over a shape with unknown rank.

  #12 opened 4 years ago by lmg1999
  12

• LVIS数据集使用coco.py问题

  #16 opened 4 years ago by xdnjust
  1

• 数据集问题

  #15 opened 4 years ago by xdnjust
  1

• 关于自己数据集里mask文件夹内图片的读取

  #14 opened 4 years ago by yanyu-holo
  3

• Open Error! Try again!

  #10 opened 4 years ago by qiqingwang577
  4

• 训练过程中出现loss为nan

  #4 opened 4 years ago by FortuneStar
  2

• ValueError: Error when checking input: expected input_image_meta to have shape (13,) but got array with shape (16,)

  #11 opened 4 years ago by qiqingwang577
  2

• 请问这个报错怎么办，是我版本不对吗? ValueError: Cannot iterate over a shape with unknown rank.

  #8 opened 4 years ago by qiqingwang577
  4

• predict.py

  #9 opened 4 years ago by babyta
  1