当前位置:网站首页>ECCV 2020 double champion team, take you to conquer target detection on the 7th
ECCV 2020 double champion team, take you to conquer target detection on the 7th
2022-06-26 05:05:00 【Paddlepaddle】
What is the hottest direction of computer vision ?
Target detection, of course !
object detection , It is one of the core problems in the field of computer vision , In the past two years, there have been hundreds of relevant papers on the global summit , Attracted more and more attention . Whether it's face recognition 、 Autopilot 、 Text detection 、 human-computer interaction , Are inseparable from the cornerstone of target detection .

With the improvement of accuracy year by year , The landing of target detection technology is becoming more and more mature , It has been widely used in industry 、 traffic 、 remote sensing 、 Medical care 、 Entertainment and so on !

Baidu as 「 China AI Head goose 」 In the research of target detection algorithm 、 Industrial practice 、 Industrial landing , Accumulated rich experience !
2020 year 7 month , Baidu PaddleDetection The R & D team explored and opened source Industrial target detection model PP-YOLO, The accuracy and prediction speed are better than YOLOv4, Be in the leading position in the industry !
2020 year 8 month , Baidu vision team is at the top level in the international vision field ECCV 2020 Excellent performance in , Take the field of target detection Tiny Object Detection and COCO Of “ Double champion ”!
To help more developers understand the field of object detection , Baidu vision team and PaddleDetection The R & D team is strong and strong , Combined with many years of practical experience , Teach in person , Lead everyone to tackle target detection !
Download the installation command ## CPU Version installation command pip install -f https://paddlepaddle.org.cn/pip/oschina/cpu paddlepaddle## GPU Version installation command pip install -f https://paddlepaddle.org.cn/pip/oschina/gpu paddlepaddle-gpu
7 Sky conquers target detection
- Top famous teachers give each other their best
The global ECCV Target detection double champion team teaching and research ,PaddleDetection Senior R & D instructors , Based on years of top research and 100+AI Industry landing experience , Live lectures throughout the course .
- Delve into the principles of the algorithm
Take apart in detail RCNN、YOLO、AnchorFree Series algorithm , Also contains SOTA Industrial test model for performance PP-YOLO, Take you step by step to clear the fog !
- Real industrial projects
In depth analysis of 4 Big Baidu features Real industrial cases : Road sign detection 、 Industrial quality inspection 、 Power inspection 、 Intensive commodity inspection in commercial supermarkets .
- Hand to hand CVPR2020 The latest competition
2019 year CVPR World champion of target detection , First public sharing of champion's Secret script , make a show of combat 2020 The latest competition in , be based on COCO Dense object detection in data sets !
Special Gospel
Excellent students have the opportunity to join Baidu
Graduation mail the paper certificate of graduation
In depth communication with Baidu senior R & D personnel throughout the whole process
Outstanding contributors to open source ecology , Give priority to promotion to PPDE( Flying propeller technician )
The whole course is free , Including live broadcast 、 practice 、 Homework correction 、 Answer questions, etc , Free of charge GPU resources
Syllabus
01 Overview of target detection tasks
1. The development of target detection 、 Important basic concepts (bbox、nms、IOU、mAP)
2. Introduction to dataset format
3. PaddleDetection Overall introduction and use
【 actual combat 】 Road sign detection
02RCNN Detailed explanation of a series of target detection algorithms
1. RCNN Introduction to the series of models
2. FasterRCNN Detailed explanation of the module
3. PaddleDetection Training RCNN Model
03RCNN A series of algorithm optimization strategies and cases
1. FPN Introduce
2. CasacdeRCNN Introduce
3. Optimization of speed and accuracy of two-stage model
4. Industrial quality inspection cases
【 actual combat 】 be based on RCNN Industrial quality inspection defect detection
04YOLO Detailed explanation of a series of target detection algorithms
1. YOLO Introduction to the series
2. YOLOv3 Detailed explanation of the module
3. Training YOLO Model
05PP-YOLO Optimization strategy and case sharing
1. PP-YOLO Introduce
2. YOLO Optimization of model speed and accuracy
3. Power inspection cases
【 actual combat 】PP-YOLO Industrial quality inspection defect detection
06AnchorFree Introduction and practice of a series of algorithms
1. AnchorFree Series introduction
2. FCOS Introduce
3. CornetNet/TTF-Net Introduce
【 actual combat 】AnchorFree Industrial quality inspection defect detection
07 The world champion will take the top hand in hand
1. Release 2020 year CVPR The latest competition question
2. World champions share 2019 Game script
【 actual combat 】 Intensive commodity inspection in commercial supermarkets
( Dense object detection )
About Instructor

Qingqing teacher : Baidu senior R & D Engineer
be responsible for Flying propeller Visual direction , Years of experience Flying propeller frame 、 Visual algorithm development , from 0 To 1 Participated in Flying propeller System and visual model library development , In charge of Flying propeller classification 、 Division 、 object detection 、 Generation and other model development , And the development of model compression tools , Rich experience in system optimization and visual algorithm application .

Teacher pengpeng : Baidu senior R & D Engineer
Server side PP-YOLO The main author of the model ,PaddleDetection The core R & D backbone of the target detection library , Yes YOLO Have a deep understanding and Research on the principle and optimization of series models , Develop and publish a variety of server-side and mobile side YOLOv3 Model , It has reached the forefront of the industry in terms of precision and speed .

Guanguan teacher : Baidu senior R & D Engineer
Flying propeller Target detection kit PaddleDetection Core R & D backbone , from 0 To 1 Created Flying propeller End to end development kit for target detection , The effect of the developed detection model has reached the forefront of the industry , With years of algorithm accumulation and practical experience in target detection direction , Have a deep understanding of the field of target detection .
Learning motivation
Worry about not being able to keep learning ? Don't worry , We need encouragement on the way to growth .
We set up a wonderful game PK link , We are ready to 100 A beautiful gift is waiting for you !

Baidu Flying propeller Official certificate of completion
Sweep yards attention 【 Flying propeller 】 official account
reply : The champion team
Get exclusive registration channel


Download the installation command ## CPU Version installation command pip install -f https://paddlepaddle.org.cn/pip/oschina/cpu paddlepaddle## GPU Version installation command pip install -f https://paddlepaddle.org.cn/pip/oschina/gpu paddlepaddle-gpu
边栏推荐
- 2022.1.24
- 天才制造者:独行侠、科技巨头和AI|深度学习崛起十年
- First day of deep learning and tensorflow learning
- Rsync common error messages (common errors on the window)
- UWB超高精度定位系统架构图
- Lstms in tensorflow_ Cell actual combat
- Dbeaver installation and configuration of offline driver
- PHP one sentence Trojan horse
- 2022.2.11
- 6.1 - 6.2 Introduction à la cryptographie à clé publique
猜你喜欢

2.22.2.14

Record a circular reference problem

Mise en œuvre du routage dynamique par zuul
![C# 39. string类型和byte[]类型相互转换(实测)](/img/33/046aef4e0c1d7c0c0d60c28e707546.png)
C# 39. string类型和byte[]类型相互转换(实测)

Codeforces Round #800 (Div. 2)

AD教程系列 | 4 - 创建集成库文件

Multipass Chinese document - remote use of multipass

ModuleNotFoundError: No module named ‘numpy‘

86. (cesium chapter) cesium overlay surface receiving shadow effect (gltf model)

UWB超高精度定位系统原理图
随机推荐
Collections and dictionaries
Rsync common error messages (common errors on the window)
[ide (imagebed)]picgo+typora+aliyunoss deployment blog Gallery (2022.6)
【Latex】错误类型总结(持更)
Zuul 实现动态路由
ROS notes (07) - Implementation of client and server
Create a binary response variable using the cut sub box operation
[greedy college] Figure neural network advanced training camp
C# 40. byte[]与16进制string互转
date_ Range creation date range freq parameter value table and creation example
[quartz] read configuration from database to realize dynamic timing task
DBeaver 安装及配置离线驱动
Day4 branch and loop jobs
Astype conversion data type
dijkstra
2022.2.16
Method of saving pictures in wechat applet
2022.2.10
LeetCode 19. 删除链表的倒数第 N 个结点
Machine learning final exercises