1.Retail intelligent video surveillance -- Pedestrian ReID, commodity recognition
No one container
Intelligent video monitoring assists stores and enterprises to manage goods and analyze consumer behavior.Computer vision technologies such as pedestrian ReID and commodity recognition are applied to retrieve pedestrian images across devices according to image and video data, detect and identify commodity information, and analyze user consumption behavior and consumption trend.Due to the difference between different camera equipment, as well as different wearing, Angle, shelter, posture, etc., the recognition accuracy of the algorithm is greatly affected.Dust with data acquisition and tagging integration services, according to different scenarios, such as unmanned container, convenient business super, clothing stores and pharmacies), different data types (such as images and video), thousands of commodity labeling of acquisition and recognition, object behavior and attitude analysis, generated a large number of structured data with high quality, effectively improve the robustness and adaptability of the model.
2.Intelligent customer service robot
Accurately identify and capture user intentions through multiple rounds of human-computer interactive voice and text dialogues.Voice recognition and natural language processing technologies are used to transcribe speech to text, extract keywords and entities, read and understand users' questions, capture effective information such as commodity introduction, logistics inquiry, promotion activities, maintenance progress, etc., analyze instructions and make intelligent answers.Dust in the vertical electric commercial retail, has many scene project marked with experience, for the ambiguous dialogue scenes grinding out a set of effective labelling specifications, to accumulate a batch of experienced marking and quality inspection team, at the same time have a auxiliary tagging system automation algorithm, simplify and automate common q&a chat scene.All-weather intelligent customer service robots can save labor costs and improve business efficiency, and help enterprises train more mature algorithm models of customer service robots.
