JDD Silicon Valley AI Lab

JD Digits AI Lab is located at the heart of the Silicon Valley. Our research team consists of dozens of world-class AI experts. We work on cutting edge AI technologies and their real-world applications, specifically in the areas of computer vision, natural language processing, speech recognition, and machine learning. Our mission is to build the most-intelligent AI engines that revolutionize our products and business for better serving hundreds of millions of JD Digits customers.

Our computer vison team is developing innovative solutions for challenging real-world problems. The main areas of research are face analysis, human body analysis, text recognition, image and video understanding, 3D vision, and robot vision. We have built a number of world-class computer vision systems for our business needs. For example, our face recognition and face anti-spoofing technologies provide seamless identify verification in both online and offline scenarios, and have been widely applied to financial products, digital marking products, and intelligent city products. Our 3D vision technologies are able to identify pigs, count pigs, estimate pigs' weight, and monitor pigs' health and body conditions, which significantly improves operational efficiency and disease control of pig farming.

Our natural language processing team is developing customer service bots, sales bots, and personal finance assistant to make our customers more efficient and effective to purchase our products and services. Our chatbots are aimed to logically answer questions from our customers and intelligently guide them to purchase the needed products and services. To achieve this goal, we are diving into cutting edge natural language processing algorithms for dialogue systems based on deep learning and knowledge graph.

Our speech team carries out applied research in two major directions: speech recognition and text-to-speech synthesis. These two directions have various dialogue-based application scenarios, such as app-based chatbots and automatic call center. Up till now, our team has launched multiple speech-centered products and continues to evolve the algorithms to enhance the quality of speech services, and consequently, to improve customer experience of speech-related products.

Our machine learning team is working on two major topics: deep user understanding and deep item understanding. To capture rich user interaction information such as purchasing, browsing, searching in both e-commerce and customer finance domains, we fuse heterogeneous data along time series, and learn a high-dimensional user embedding via the state-of-the-art deep learning algorithms. We also build knowledge graphs and innovate graph embedding algorithms to build item/product embeddings. These two core technologies have been widely applied in the areas of user growth and personal recommendation scenarios.

Computer Vision
Face Analysis
Face detection
Facial landmark detection
Face recognition
3D face reconstruction
3D face recognition
Human Body Analysis
Human detection
Human pose estimation
Human body tracking
Human activity recognition
Human shape estimation
Text Recognition
ID card recognition
Bank card recognition
Driver license card recognition
Corporate license card recognition
Receipt recognition
Generic text recognition
Image and Video Understanding
Image Review
Product Recognition
QR code recognition
Vehicle damage estimation
Computer Vision for Pig Farming
Pig face recognition
Pig counting
Pig weight estimation
Pig health monitoring
3D pig reconstruction
Pig shape evaluation
Robot Vision
3D mapping
Semantic segmentation
Moving object tracking
Object motion estimation
Motion planning
Natural Language Processing
Text Understanding
User intent/topic classification
User sentiment analysis
Text representation and semantic matching
Context understanding and multi-turn dialogue
Named entity recognition
Knowledge graph
Text Synthesis
Text generation
FAQ generation
Speech Recognition
Near-field speech recognition
Far-field speech recognition
Audio transcription
Long speech recognition
Real-time speech recognition
Speech Synthesis
Customized pronunciation
Multi-speaker pronunciation
Machine Learning/Deep Learning
User embedding
Product embedding
Graph embedding
User growth
Personalized recommendation
Automatic machine learning
User behavior recognition
Unsupervised learning
Transfer learning
Semi-supervised learning
There's more to JD Digits
Digital Finance
Intelligent Cities
Digital Agri-husbandry
Digital Marketing
Digital Asset Management