The Baidu Shanxi Data Annotation Base needed a scalable way to collect and label training data for pandemic-era AI projects such as CT-assisted virus screening and masked-face recognition, while keeping a large workforce productive through remote operations.
Reported impact
The base helped support pandemic-response AI deployments, reached 99% masked-face recognition accuracy, enabled more than 1,300 people to resume work online, delivered re-employment for 2,000+ annotators, and underpinned the landing of 35 AI enterprises in Shanxi.
Baidu AI Cloud's Shanxi Data Annotation Base used Baidu's Data Annotation Service & Content Review Service together with its Data Collection and Annotation Solution to run a human-in-the-loop training-data pipeline for epidemic-response and broader AI programs. Source data such as lung CT scans and masked-face photos were gathered into a cloud-based annotation environment, where trained annotators worked remotely to label viral regions in medical images and fine facial landmarks such as eyebrows, glasses, and cheekbones. The collected labels were reviewed and organized as structured training data for downstream algorithms, improving CT pre-screening and masked-face recognition accuracy. The same operating model also supported workloads in autonomous driving, image classification, and speech recognition by combining online task execution, workforce training, and large-scale distributed annotation.
Lung CT images and masked-face photos are collected into Baidu's cloud annotation environment. Annotators working remotely label viral image regions and facial keypoints such as eyebrows, glasses, and cheekbones, then reviewed labels are delivered to downstream AI algorithms that support CT pre-screening, face-gate access, and multi-person temperature measurement deployments.
After collecting a large number of masked-face photos, annotators precisely labeled eyebrows, glasses, cheekbones, and other facial keypoints, and more labeled feature points let AI recognize faces under mask-wearing scenarios more accurately (original: 在采集大量的戴口罩的人脸照片后，数据标注师对人脸的眉毛、眼镜、颧骨等人脸关键点进行精准的标注，标注的特征点越多，AI就越能精确地识别戴口罩场景下的人脸。)
As of now, the Baidu Shanxi Data Annotation Base has become China's number-one standalone data-annotation base by headcount and output value, with more than 2,000 AI data annotators, revenue above RMB 100 million, and 35 resident enterprises (original: 截止目前，百度山西数据标注基地已经成为中国人员和产值规模第一的单体数据标注基地。AI数据标注师从业人员超过2000人，实现营业收入超亿元，企业入驻35家。)
Baidu AI multi-person temperature measurement has been deployed in hundreds of public places nationwide and detected 27 million person-visits within two months (original: 目前，百度AI多人测温已在全国百处公共场所落地部署，落地两个月内就累计检测2700万人次。)
Data annotators at the Baidu Shanxi Data Annotation Base teach AI to recognize data, and enough high-quality data lets AI perceive, think, and make decisions more like humans (original: 百度山西数据标注基地数据标注师的工作就是教会AI认识数据,有了足够多、足够好的数据,AI才能学会像人一样去感知、思考和决策,更好地为人类服务。)
During the pandemic, based on Baidu's cloud-based annotation and remote work, more than 1,300 people at the base resumed work online and supported multiple technology-for-epidemic-prevention projects nationwide (original: 疫情期间，基于百度的云上标注和远程办公，百度山西数据标注基地1300多人线上复工，为全国各地多个“科技防疫”项目的实施应用提供了重要支撑。)
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
Baidu AI Cloud Sample Dec 29, 2025 Beijing Yilian Tongda Technology Co., Ltd. Yilian Tongda needed a safer and more efficient way to help power-industry staff work with scattered standards, maintenance records, inspection documents and risk-control knowledge that were difficult to search and keep… Agent Technology · Qianfan AppBuilder · Qianfan ModelBuilder Baidu AI Cloud Sample Dec 29, 2025 Hualvwang Hualvwang needed a way to help ordinary users describe legal issues more precisely, retrieve relevant legal knowledge faster and scale personalized legal-service triage across a very large user base. Daily ERNIE invocation peaks exceeded 10 million. Qianfan AppBuilder · Qianfan ModelBuilder · Recommended Questions Baidu AI Cloud Sample Dec 29, 2025 Shanghai Xinzhi Software Co., Ltd. (上海新致软件股份有限公司) Shanghai Xinzhi Software needed a scalable way to help automotive sales teams cope with high staff turnover, expensive talk-track training, fragmented customer-profile information, and weak use of historical call data… Average sales score increased 25%. Baidu Search plugin · ERNIE-4.0-Turbo · Qianfan AppBuilder 