TPU processor, 16 channels HD video intelligent analysis, 32 channels of full HD video decoding
TPU processor, 32 channels HD video intelligent analysis, 32 channels of full HD video decoding, 12 channels of full HD video encoding
Based on the RISC-V core, operating at a frequency of 2GHz, the processor features a single SOC with 64 cores and 64MB shared L3 cache.
Based on RISC-V 3-5M slightly intelligent deep learning vision processor
Based on RISC-V 5M light intelligent deep learning vision processor
RISC-V 5M light intelligent deep learning vision processor
5M light intelligent deep learning vision processor
4K Super Definition Deep learning vision processor
5M high-performance Deep learning vision processor
5M light intelligent Deep learning vision processor
960-channel HD video decoding, 480-channel HD video analysis
576-channel HD video decoding, 288-channel HD video analysis
BM1684X, 416-channel HD video analysis
X86 host processor,288-channel HD video analysis
BM1684X, 32-channel HD Video Analysis
BM1684, 16-Channel HD Video Analysis
BM1684, 192-channel HD video analysis
BM1684, 8-channel HD video analysis
CV186AH, 8-channel HD Video Analysis
BM1688, 16-channel HD Video Analysis
72-channel HD video decoding, 72-channel HD video analysis
96-channel HD video decoding,48-channel HD video analysis
32-channel HD video decoding,16-channel HD video analysis
32-channel HD video decoding, 32-channel HD video analysis
32-channel HD video decoding, 32-channel HD video analysis
32-channel HD video decoding, 16-channel HD video analysis
32-channel HD video decoding, 16-channel HD video analysis
Deep Learning Developer Product Portfolio
Deep learning intelligent analysis helps make city management more efficient and precise
Using deep learning video technology to analyze sources of dust generation and dust events, contributing to ecological environmental protection
Empower prison management with intelligent monitoring of key controlled areas through smart video analysis
Using deep learning intelligent analysis to monitor scenarios such as safety production, urban firefighting, and unexpected incidents for emergency regulation.
Using specific deep learning algorithms to watermark, blur, or apply other methods to streaming videos, achieving video confidentiality and preventing leaks
SOPHGO with the SOPHON.TEAM ecosystem to build a data intelligence content governance solution.
Using deep learning technology to detect and analyze individuals, vehicles, and security incidents in grassroots governance
Real-time compression and transcoding of video to the cloud and monitoring of abnormal events, enhancing the ability to detect and handle road safety incidents
Empowering the problems of traffic congestion, driving safety, vehicle violations, and road pollution control
Utilizing domestically developed computational power to support the structured analysis of massive volumes of videos, catering to practical applications in law enforcement
Build a "smart, collaborative, efficient, innovative" gait recognition big data analysis system centered around data
To rapidly construct business capabilities that integrate multidimensional data including people, vehicles, and traffic flow for users
Effectively resolving incidents of objects thrown from height, achieving real-time monitoring of such incidents, pinpointing the location of the thrown object, triggering alerts, and effectively safeguarding the safety of the public from falling objects
Using edge computing architecture to timely and accurately monitor community emergencies and safety hazards
SOPHGO with SOPHON.TEAM ecosystem partners to build a deep learning supervision solution for smart hospitals, enhancing safety management efficiency in hospitals
SOPHGO with SOPHON.TEAM ecosystem partners to build a smart safe campus solution
Using a combination of cloud-edge deep learning methods to address food safety supervision requirements across multiple restaurant establishments, creating a closed-loop supervision system for government and enterprise-level stakeholders
Providing deep learning capabilities for the financial, insurance, and various business service industries to enhance operational efficiency and improve service quality
SOPHGO with SOPHON.TEAM ecosystem partners to offer a "Deep Learning Video Analysis + Restaurant Front-of-House Management" solution
SOPHON's self-developed computing hardware devices, such as SG6/SE5/SE6, equipped with SOPHON.TEAM video analysis algorithms, are used to make industrial safety production become smarter
Provided safety monitoring solutions for violations and abnormal events in offices, quality inspection, weighing rooms, storage areas and other areas of large storage parks such as granaries and cotton warehouses
SOPHON.TEAM is collaborating with ecological partners to develop a comprehensive solution for ensuring the safety of tobacco industry production and control
In collaboration with SOPHON.TEAM and its ecological partners, SOPHGO utilizes domestically developed computing power products as the hardware foundation to build a safety production management system and improve the safety production management level of liquor enterprises
Combining deep learning, edge computing and other technologies, it has the ability to intelligently identify people, objects, things and their specific behaviors in the refueling area and unloading area. It also automatically detects and captures illegal incidents at gas stations to facilitate effective traceability afterwards and provide data for safety management.
SOPHGO, in collaboration with SOPHON.TEAM and its ecosystem partners, is focusing on three major scene requirements: "Production Safety Supervision," "Comprehensive Park Management," and "Personnel Safety & Behavioral Standard Supervision." Together, they are developing a comprehensive deep learning scenario solution, integrating "algorithm + computing power + platform."
SOPHGO, cooperates with SOPHON.TEAM ecological partners to build a deep learning monitoring solution for safety risks in chemical industry parks
SOPHGO with SOPHON.TEAM ecosystem partners to build a Smart Computing Center solution, establishing a unified management and scheduling cloud-edge collaborative smart computing center
SOPHGO, in collaboration with SOPHON.TEAM ecosystem, have jointly developed a set of hardware leveraging domestically-produced deep learning computational power products. This is based on an AutoML zero-code automated deep learning training platform, enabling rapid and efficient implementation of deep learning engineering solutions
本次赛题提供的数据集包含image和label两个文件夹。image文件夹由2885张训练集和133张测试集组成;label文件夹中包含2885个带有对应每张图片全部标签的txt文件。标签共包括三种类别:机动车、行人和非机动车。数据标签格式如下:
<class_id> <ct_x> <ct_y> <w> <h>
标签对应id:0:car, 1:person, 2:bike
E.g.
0 0.4284 0.6977 0.2984 0.5361
1 0.2784 0.6125 0.0380 0.0361
2 0.6706 0.7072 0.0255 0.0645
参赛者在训练完成并对测试集图片进行推理后,将测试集图片的检测结果以与训练集标签同名的txt文件保存,并上传至result_submit/文件夹下,具体保存方式见result_submit/README.md。上传格式实例如下所示:
<class_name> <confidence> <left> <top> <right> <bottom>
E.g.
car 0.399786 3 642 125 690
person 0.395193 1549 621 1686 796
bike 0.386811 373 647 395 701
P.S. 我们在scripts/文件夹中为选手提供脚本,可将<class_id> <confidence> <ct_x> <ct_y> <w> <h>
格式转换为提交所需的 <class_name> <confidence> <left> <top> <right> <bottom>
格式。
将检测结果保存至input/detection-results
将测试图片保存至images/
运行如下指令
python3 convert_gt_yolo.py
选手提交成绩后,若审核无误将会每日于result_submit/readme.md中更新分数,每周于赛事官网页面更新分数
通过mAP(mean Average Precison)指标评测模型精度。mAP即各类别APIoU=0.5的平均值,其中APIoU=0.5为IoU阈值为0.5 的平均精度。初赛最终得分计算公式为:score=mAP*100,分数高者为优。
1. 通过与初赛相同的mAP(mean Average Precison)指标评测模型精度。
2. 通过模型推理时间i_time评测模型性能,i_time为测试集图片推理的平均时间,单位为ms。
3. 最终得分计算公式为:score=mAP*100+(1000-i_time)*0.1,分数高者为优。