Mask_RCNN:ValueError: Dimension 1 in both shapes must be equal, but are 8 and 324.

编程语言 2018-06-04 05:09:30 阅读次数: 0

前几天有人在之前Mask_RCNN预测文章中：

https://blog.csdn.net/qq_15969343/article/details/80388311

评论了该问题：

我看了一下，主要是因为直接使用了官方给出的COCO权重进行了预测，但是可能COCO的80类物体中存在有想预测的东西，因此直接把NUM_CLASS改为了1+1；这样显然是不对的，因为1个类占了4个通道，因此是（1+80）*4=324，但是现在改为了（1+1）*4=8【8！=324】，模型自然会报错！

因此使用官方给出的权重，而不是自己训练的，同时遇到这个问题的同学，请把这个地方改为：1+80：

当然，我们可以验证一下：将1+80改为1+70，会出现以下错误：

（1+70）*4正好等于284！结论正确！

猜你喜欢

转载自blog.csdn.net/qq_15969343/article/details/80559154

Mask_RCNN:ValueError: Dimension 1 in both shapes must be equal, but are 8 and 324.

ValueError: Dimension 1 in both shapes must be equal, but are 8 and 324 for 'Assign_376' (op: 'Assig

【Python】ValueError: x and y must have same first dimension, but have shapes (5,) and (4,)

ValueError: Shape must be rank 2 but is rank 0 for 'MatMul_8' (op: 'MatMul') with input shapes: [],

【TensorFlow】ValueError: Shape must be rank 1 but is rank 0 for ' ’ with input shapes: [].问题

Tensorflow报错：ValueError: Stride must be > 0, but got 0 for '...' with input shapes: [...], [...]

ValueError: Shape must be rank 1 but is rank 0 for 'bn_conv1/cond/Reshape_4' (op: 'Reshape') with input shapes: [1,64,1,1], [].

RuntimeError: Sizes of tensors must match except in dimension 1. Got 61 and 60 in dimension 2

BDD数据集（mask_rcnn）1

ValueError: Must have equal len keys and value when setting with an iterable

Tensorflow报错: Shape must be rank 2 but is rank 1 for 'MatMul' (op: 'MatMul') with input shapes

【Bug记录-1】Androidstudio升级至3.1报错All flavors must now belong to a named flavor dimension

The expanded size of the tensor (256) must match the existing size (81) at non-singleton dimension1

RuntimeError: invalid argument 0: Sizes of tensors must match except in dimension 1 #805

RuntimeError: invalid argument 0: Sizes of tensors must match except in dimension 1. Got 14 and 13

Sizes of tensors must match except in dimension 1. Expected size 24 but got size 25 for tensor numbe

记录解决RuntimeError: Sizes of tensors must match except in dimension 1. Expected size 27 but got size

mask_rcnn keras源码跟读1）模型搭建

Assign requires shapes of both tensors to match. lhs shape= [3,3,256,8] rhs shape= [3,3,256,84]

InvalidArgumentError (see above for traceback): Assign requires shapes of both tensors to match

tensorflow error : Assign requires shapes of both tensors to match.

Shape must be rank 1 but is rank 0 for 'bn1/cond/Reshape_4' (op: 'Reshape') with input shapes: [1,64

Keras：ValueError: Negative dimension size caused by subtracting 2 from 1 for

《机器学习实战》第8章 ValueError: Masked arrays must be 1-D

svn both sides of the move must be committed together

TypeError: float() argument must be a string or a number, not 'Dimension'

ValueError: operands could not be broadcast together with shapes (9,8) (8,9)

scipy interpolation ValueError: x and y arrays must be equal in length along interpolation axis

ValueError: Dimensions must be equal, but are 16 and 28 for 'seq2seq/decoder/train/decoder

关于李沐深度学习The size of tensor a (10) must match the size of tensor b (256) at non-singleton dimension 1

今日推荐

周排行

深度学习------Lingvo框架下的加速通道GPipe

webjars管理静态资源

C专家编程_2.2

mysql 源码安装

json文件操作

123231432

注解的实现

Spring MVC 控制器

《人月神话》读后感二

C#使用HttpWebRequest和HttpWebResponse上传文件示例

每日归档

2024-09-08(0)

2024-09-07(0)

2024-09-06(0)

2024-09-05(0)

2024-09-04(0)

2024-09-03(0)

2024-09-02(0)

2024-09-01(0)

2024-08-31(0)

2024-08-30(0)