Committed to providing users with the most accurate speech recognition and other technology services, and let man and machine communicate in the most natural and convenient way.
Committed to letting the machine imitate human 'thinking', understand what you say and tell what you want to know.
Committed to using the text recognition, face recognition, content understanding and augmented reality technologies and let the machine perceive the world through vision.
Focused on the theme of clustering analysis, information classification, non-structured knowledge extraction, text understanding and hotspot tracking technology.
Amber is a distributed computing platform with a set of API and libraries, and is used to support WeChat AI research on machine learning and deep learning.
We propose a novel, efficient approach for distributed sparse learning with observations randomly partitioned across machines. In each round of the proposed method, worker machines compute the gradient of the loss on local data and the master machine solves a shifted l₁ regularized loss minimization problem. After a number of communication rounds that scales only logarithmically with the number of machines, and independent of other parameters of the problem, the proposed approach provably matches the estimation error bound of centralized methods.
ICML · 2017
To the same utterance, people’s responses in everyday dialogue may be diverse largely in terms of content semantics, speaking styles, communication intentions and so on. Previous generative conversational models ignore these 1-to-n relationships between a post to its diverse responses, and tend to return high-frequency but meaningless responses.
AAAI · 2017
How do social groups, such as Facebook groups and Wechat groups, dynamically evolve over time? How do people join the social groups, uniformly or with burst? What is the pattern of people quitting from groups? Is there a simple universal model to depict the come-and-go patterns of various groups? In this paper, we examine temporal evolution patterns of more than 100 thousands social groups with more than 10 million users. We surprisingly find that the evolution patterns of real social groups goes far beyond the classic dynamic models like SI and SIR. For example, we observe both diffusion and non-diffusion mechanism in the group joining process, and power-law decay in group quitting process, rather than exponential decay as expected in SIR model. Therefore we propose a new model comeNgo, a concise yet flexible dynamic model for group evolution. Our model has the following advantages: (a) unification power: it generalizes earlier theoretical models and different joining and quitting mechanisms we find from observation. (b) succinctness and interpretability: it contains only six parameters with clear physical meanings. (c) accuracy: it can capture various kinds of group evolution patterns preciously and the goodness of fit increase by 58% over baseline. (d) usefulness: it can be used in multiple application scenarios such as forecasting and pattern discovery.
KDD · 2016
You can send voice messages or text messages to friends via voice input
WeChat - Contacts - Choice - Messages
You just to gently sweep, could complete the payment, subscription, download software and a series of needs
WeChat - Discover - Scan - QR Code / Cover
You can look at the words you want to know, whether it is a handwritten letter or an original magazine
WeChat - Discover - Scan - Translation
It will joke, confabulate, give you ideas, and teach you great wisdom
WeChat - Contacts - New Friends - WeChat ID
You can learn more about the Music / TV Show information by Shake the Music / TV
WeChat - Discover - Shake - Music / TV
You can use unique voice for data information to form a natural encryption protection. Security is so self-willed
Me - Settings - Account Security - Voiceprint