当前位置:网站首页>[PaddleSeg 源码阅读] PaddleSeg计算Dice
[PaddleSeg 源码阅读] PaddleSeg计算Dice
2022-07-04 03:32:00 【氵文大师】
D i c e Dice Dice 系数是一种集合相似度量函数,通常用于计算两个集合的相似度,取值范围是 [ 0 , 1 ] [0, 1] [0,1]:
D i c e ( X , Y ) = 2 ∣ X ∩ Y ∣ ∣ X ∣ + ∣ Y ∣ Dice(X, Y) = \frac {2 | X\cap Y|} { |X| + |Y| } Dice(X,Y)=∣X∣+∣Y∣2∣X∩Y∣
∣ . ∣ |.| ∣.∣ 表示取集合元素的数量函数
D i c e ( X , Y ) = 两 者 交 集 元 素 的 个 数 × 2 两 者 元 素 数 量 之 和 Dice(X, Y) = \frac {两者交集元素的个数 \times 2} {两者元素数量之和} Dice(X,Y)=两者元素数量之和两者交集元素的个数×2
PaddleSeg 中的实现:
def dice(intersect_area, pred_area, label_area):
""" Calculate DICE. 计算 DICE 值. Args: intersect_area (Tensor): The intersection area of prediction and ground truth on all classes. pred_area (Tensor): The prediction area on all classes. label_area (Tensor): The ground truth area on all classes. Returns: np.ndarray: DICE on all classes. float: mean DICE of all classes. """
# 将三者转化为 np.adarray
intersect_area = intersect_area.numpy()
pred_area = pred_area.numpy()
label_area = label_area.numpy()
# 两个集合元素的数量
union = pred_area + label_area
class_dice = []
for i in range(len(intersect_area)): # 在每一类中开始迭代
if union[i] == 0:
# 如果分子为 0,也就是当前图片没有该类元素
dice = 0
else:
dice = (2 * intersect_area[i]) / union[i]
class_dice.append(dice)
mdice = np.mean(class_dice)
return np.array(class_dice), mdice
intersect_area, pred_area, label_area 这三个参数是通过:
calculate_area(pred, label, num_classes, ignore_index=255)
来计算的,接上一篇博客:
[PaddleSeg 源码阅读] PaddleSeg计算 mIoU
边栏推荐
- Apple submitted the new MAC model to the regulatory database before the spring conference
- Stm32bug [the project references devices, files or libraries that are not installed appear in keilmdk]
- Solve the problem that the tabbar navigation at the bottom of vantui does not correspond to the page (window.loading.hash)
- Li Chuang EDA learning notes 13: electrical network for drawing schematic diagram
- 7 * 24-hour business without interruption! Practice of applying multiple live landing in rookie villages
- JSON string conversion in unity
- Love and self-discipline and strive to live a core life
- Ningde times and BYD have refuted rumors one after another. Why does someone always want to harm domestic brands?
- JS object definition
- false sharing
猜你喜欢
![[latex] production of complex tables: excel2latex and detail adjustment](/img/39/0d448ddf006eda262de3ed75666354.jpg)
[latex] production of complex tables: excel2latex and detail adjustment

Is it really so difficult to learn redis? Today, a fan will share his personal learning materials!

WP collection plug-in free WordPress collection hang up plug-in

Have you entered the workplace since the first 00???

150 ppt! The most complete "fair perception machine learning and data mining" tutorial, Dr. AIST Toshihiro kamishima, Japan

Third party login initial version

Summary of Chinese remainder theorem

Consul of distributed service registration discovery and unified configuration management

Teach you how to optimize SQL

7 * 24-hour business without interruption! Practice of applying multiple live landing in rookie villages
随机推荐
If you have just joined a new company, don't be fired because of your mistakes
Webhook triggers Jenkins for sonar detection
96% of the collected traffic is prevented by bubble mart of cloud hosting
Package and download 10 sets of Apple CMS templates / download the source code of Apple CMS video and film website
Monitoring - Prometheus introduction
Leetcode 110 balanced binary tree
Osnabrueck University | overview of specific architectures in the field of reinforcement learning
Safety tips - seat belt suddenly fails to pull? High speed police remind you how to use safety belts in a standardized way
system information
warning: LF will be replaced by CRLF in XXXXXX
Sword finger offer:55 - I. depth of binary tree
Contest3145 - the 37th game of 2021 freshman individual training match_ D: Ranking
XSS prevention
Leecode 122. Zuijia timing of buying and selling stocks ②
Consul of distributed service registration discovery and unified configuration management
CSP drawing
Contest3145 - the 37th game of 2021 freshman individual training match_ E: Eat watermelon
Which product is better if you want to go abroad to insure Xinguan?
No clue about the data analysis report? After reading this introduction of smartbi, you will understand!
Base d'apprentissage de la machine: sélection de fonctionnalités avec lasso