Dear data scientists !
Have our preheating game Unlocked
been completed?
has fully mastered the competition skills and is ready to fight at any time?
"Combining the power of Z to unlock the extreme wisdom realm"
HP Z series algorithm competition
September 29th has sailed to sail
official competition questions have been released
invite you who love data science
and HP
unlock the extreme wisdom realm
competition questions list
[Task Introduction]
Object detection As one of the hot technologies for the current landing of computer vision , it has been widely used in various scenarios such as autonomous driving, smart parks, industrial detection and satellite/drone remote sensing. This competition task is target detection with angle information, combined with a large amount of high-resolution visible vehicle target data as an auxiliary to achieve the extraction of the position, direction and category information of vehicle targets in visible and infrared images.
Figure 1-1 Schematic of the target detection result with angle information
left: visible light detection result, right: infrared detection result
[Dataset Information]
This competition dataset is mainly divided into two parts. Some of are high-definition, visible light traditional rectangular vehicle target image data. is used as auxiliary data for the competition task, which can be used for large-scale data pre-training and other related strategies to provide assistance for target detection in subsequent small samples; the other part of is aligned visible light and infrared rotating vehicle target image data. is used as the main data for the competition task, which is used for the participating team to complete the intelligent model of visible light and infrared rotating vehicle target detection based on a small amount of paired data, and evaluate it in the remaining data.
HD resolution visible vehicle image data schematic diagram
Alignment visible light-infrared vehicle image data
On: Daytime environment, under: Night environment
[Selecting rules]
According to the overlap ratio RIoU between the predicted target box and the real box, exceeding the specified ratio is considered to be successful in detection.
match score calculation method:
mAP@RIoU=0.5:0.05:0.95.
* Detailed image data, please go to the competition official website to learn more about
See this competition question that is closely related to daily life
. Are you unable to hold back?
This data science competition is open to the whole society
Individuals, universities, scientific research units, enterprises,
Maker teams and other personnel can register
Whether you are a "novice" who urgently needs to test your skills or a "expert" who wants to show off your supernatural powers
This is a practical opportunity that cannot be missed
and there are rich bonuses waiting for you to win!
competition reward
competition organizing committee will select 6 excellent participating teams
split .1 million yuan bonus pool
first prize 1 first prize 30,000 yuan
second prize 2 second prize 20,000 yuan
third prize 3 third prize 3 Bonus 10,000
Schedule
Registration and model submission time
September 29th to November 24th
Results review
November 25th to December 7th
Final Released
December 8
Help data scientists win good results
HP Z series data science workstation strongly escort
HP Z series data science workstation is a professional computer product designed specifically for the field of data science and is designed for professional graphics and image processing, computing, design, and simulation. HP Z series data science workstations have the advantages of high computing power, high stability, high security, and easy deployment. They support developers to quickly and securely configure development environment , providing hardware support for R&D work in complex data computing scenarios in the AI field, and effectively meeting the needs of users and enterprises for high performance and reliable computing power.
Among them, HP Z series Z8 G4 desktop workstations and ZBook Studio G9 mobile workstations are the preferred development platforms for this competition.
01
HP Z series data science workstation
HP Z8 G4 desktop workstation
//
HP Z series Z8 G4 desktop workstation has a powerful hardware configuration, supports two new generation Intel® Xeon® processors, can accommodate 56 processor cores in one system, can support 3TB of high-speed memory, and provides high-performance GPU accelerated algorithm development tools.
The complex data simulation operation, advanced machine learning algorithm processing, massive data computing and other tasks involved in this competition are very suitable for data scientists to use the HP Z series Z8 G4 desktop workstation.
02
HP Z series mobile workstation
ZBook Studio G9
//
HP Z series ZBook Studio G9 mobile workstation combines outstanding portability and professional performance. Users can choose Intel® Core i9 CPU, as well as NVIDIA RTX A5500 or NVIDIA® GeForce RTX 3080 Ti GPU and 64 GB RAM to speed up workflows and help data scientists build complex components, render in real time or visualize data.
From 3D design to machine learning, the ZBook Studio G9 mobile workstation can provide powerful performance that meets all aspects of the problem-solving process, helping you to smoother this problem-solving process and win the game!
in the ocean of data, joining the mountain and joining the mountain and joining the
in the same stage with the experts in the data world, compete on the same stage with
in this battle, just wait for you to "join the power of Z"
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-