当前位置:网站首页>Unsupervised learning of visual features by contracting cluster assignments
Unsupervised learning of visual features by contracting cluster assignments
2022-07-07 11:17:00 【InfoQ】
- The comparative study of the gods in the twilight age
- “ The arms race ” The comparative study of the period is good .
- Will the whole imagenet Make a dictionary , Extract one from mini batch As a positive sample . And then randomly extract from it 4096 As a negative sample .
- Extract one from the data set mini batch Expand it , Using a twin network , Put the original image into a network , Put the enhanced graph into another network , Both are trained at the same time , Use one for both NCE loss perhaps infoNCE loss. A picture and its enlargement as a positive sample , The rest of the images and their extensions are taken as negative samples .
- Extract one from the data set mini batch Expand it twice , Using a twin network , Put a set of image enhancements into a network , Put another set of image enhancements into another network , Both are trained at the same time , Use one for both NCE loss perhaps infoNCE loss.
- It may repeatedly extract the same data . Although your data set has many pictures , But you may draw the same picture from it . In extreme cases , If you take a group of pictures as a positive sample , Then you take a group of pictures with the same repetition as the negative sample . That will affect the training .
- It may not be representative of the entire data set . For example, there are many kinds of animals in this data , But all you get is dogs , So the data is not representative .
- Of course, the more comprehensive the selection, the better the effect , But if you choose too many negative samples, it will cause a waste of computing resources .
- Let's start with the repeating question : Because you use the cluster center for comparison . Although it is a different cluster center , Then it is certainly impossible for him to repeat .
- Again, there is no representative problem : Clustering is to gather many pictures into different categories . Compare with the center of each category , Is absolutely representative .
- Let's talk about the waste of resources caused by too many negative samples in the past . If you want to make an analogy with many negative samples , You may need thousands of negative samples , And even so, it is only an approximation , And if you just compare with the cluster center , You can use hundreds or at most 3,000 Cluster centers , That's enough to say . Greatly reduce the consumption of computing resources .
边栏推荐
- 自动化测试框架
- Transaction rolled back because it has been marked as rollback-only解决
- Creative information was surveyed by 2 institutions: greatdb database has been deployed in 9 places
- 2021-05-21
- [untitled]
- Mpx 插件
- 分布式数据库主从配置(MySQL)
- [untitled]
- JSON format query of MySQL
- 关于SIoU《SIoU Loss: More Powerful Learning for Bounding Box Regression Zhora Gevorgyan 》的一些看法及代码实现
猜你喜欢
The use of list and Its Simulation Implementation
Ping tool ICMP message learning
IDEA快捷键大全
Deeply understand the characteristics of database transaction isolation
shardingsphere分库分表示例(逻辑表,真实表,绑定表,广播表,单表)
Force buckle 1002 Find common characters
[untitled]
seata 1.3.0 四種模式解决分布式事務(AT、TCC、SAGA、XA)
关于在云服务器上(这里用腾讯云)安装mysql8.0并使本地可以远程连接的方法
基于华为云IOT设计智能称重系统(STM32)
随机推荐
Template initial level template
网络协议 概念
简单介绍一下闭包及它的一些应用场景
從色情直播到直播電商
Go Slice 比较
什么是高内聚、低耦合?
基于华为云IOT设计智能称重系统(STM32)
POJ1821 Fence 题解报告
普通测试年薪15w,测试开发年薪30w+,二者差距在哪?
[untitled]
2021-04-08
PostgreSQL中的表复制
Laya common script commands
从色情直播到直播电商
学习笔记|数据小白使用DataEase制作数据大屏
Galaxy Kirin desktop operating system installation postgresql13 (source code installation)
Rolling puddle Uni_ App (VIII)
Go slice comparison
创意信息获2家机构调研:GreatDB 数据库已在9地部署
分布式数据库主从配置(MySQL)