About Me
Experience
Data Engineer
某公司 - - Now
职位: Data Engineer | 经验: 6 months | 项目: big data analyse system based on video cloud :\n*use gb28181/onvif/rtsp protocols to streaming from cameras ,and then extracting frame, decoding,push to kafka. \n*use flink to consume data from kafka and then transmit the image data to base64 for calling AI service(such as video structured analysis,image search, micro expression recognition , object detect ,trajectory tracking, facial/human/vehicle structured analysis, temperature and humidity/traffic flow statistical analysis and prediction,...), write the analyse result into hbase-es (two band index)\n\n AI system based on video cloud:\n*Support the integration of self-developed or third-party algorithms through algorithm SDKs and algorithm services; \n*Support application level algorithm scheduling strategies based on camera configuration;\n* Integrating multiple types of algorithms, including behavior analysis algorithms (safety helmets, smoking, wearing masks, fireworks, intrusion, making phone calls), video structuring algorithms (face structuring, human body structuring), facial recognition algorithms, license plate recognition algorithms, and human key point recognition algorithms,video quality diagnosis,video super resolution,etc. \n*Supports real-time detection, behavior analysis, facial recognition, video compression, fragment extraction Multiple business capabilities such as target recognition;\n\nBig Data Platform: \nCustomized development based on the third-party open source big data platform WeDataSphere. The basic platform consists of four levels: data exchange, data distribution, computation, and storage, focusing on the underlying data transmission, computation, and storage capabilities; The functional platform consists of three levels: platform tools, data tools, and application tools, focusing on the implementation of various functional tool needs of users. A complete big data platform technology system has been formed, providing one-stop rich data platform components and functional support. We have created a one-stop fully connected big data application development and management platform (data center). Having the technical ability to access, store, analyze, and visualize big data throughout the entire process.\nArtificial Intelligence Platform: \nBased on a third-party open-source artificial intelligence platform (Tianshu Platform), we have created a \"Fully Functional Deep Learning and Model Reasoning Platform (AI Platform)\" with technical capabilities for the entire process of data annotation, model development, model training, algorithm reasoning, and more. The entire platform consists of three subsystems: a one-stop AI model development platform, a high-performance deep learning framework, and a model refinement framework. Among them, the one-stop AI model development platform (referred to as the one-stop development platform) is aimed at the lifecycle of AI model production, providing functions including data processing, model development, model training, and model management,etc.
