| CN
EN
Provide end users with an integrated “cloud + terminal + core” automotive intelligent interactive service
— Qianxingyi
Capability Framework
行业痛点
Architecture
Core Technologies
• support natural expression
• end cloud seamless integration
• support personalized and dynamic models
• support all offline functions
• low resource consumption
• end to end technology, industry leading
• support Mandarin, Cantonese, English, dialects, etc
• support language mix and match recognition
• support listening mode and intelligent sentence breaking
• support multiple sound zones and simultaneously monitor wake-up recognition
• AEC / NR support software solution / hardware DSP
• support handcart interconnection (car play, carlife)
• support clear voice communication
• support off-line natural semantic understanding
• support environment context
• support what you see is what you say
• intelligent dynamic guidance system
• industry leading offline TTS
• can quickly synthesize personalized user voice
• can customize a variety of tone TTS (Star tone, sweet, etc.)
• support multi language and multi dialect TTS
• multi algorithm fusion semantic platform, more natural language expression
• natural and full tone, expressive
Full duplex interaction
• one wake-up, long-time interaction (120 seconds)
• small talk and encyclopedia that can be filtered according to semantics;
• organic combination of acoustic and semantic confidence
• any scenario can interrupt and initiate a new conversation
• any interface can support 'visible' and 'spoken'
• natural semantic understanding and freedom of expression
• offline can also support visible and spoken
Customer Groups
The customer relies on the supplier to complete the design, planning and production of the whole vehicle network, and needs to provide TSP, voice, ecology, operation, app and other related contents
To provide a complete voice assistant, you need to complete voice skills, UE interaction and UI. And provide some ecological content.
Output core capabilities, acoustic processing, wake-up, recognition, understanding, TTS, and support customers in secondary development
Provide a single capability at the customer's request
Advantage
• flexible provision of soft / hard solutions
• support 2-6 sound zones
• support Carplay, carlife, etc
• support natural language interaction
• voice search with navigation poi
• industry leading embedded TTS
• terminal: fast and timely update
• cloud: accurate and rich in content
• flexible configuration of solutions
• embedded solution (low resource configuration)
• cloud supports customized configuration
• hundreds of project experience
• tens of thousands of model information extraction
• accurate user portrait Dimension Extraction
• support deep and targeted customized development
• high openness and low customization threshold
Related
Realize second-level analysis, with the highest recognition rate in insurance, banking and other fields
Speaking tips and warnings, restraining non-compliant behaviors of agents
Daily processing capacity 1000 hours+, 100% full quality inspection
Support two scenarios of outbound and inbound calls, human-machine collaboration
Visualized real-time large screen based on voice big data, real-time public opinion warning
Support multi-channel teaching such as APP and mini programs, and teach speech in a scene interactive way
Deep integration of AI based on the two core capabilities of speech recognition and speech synthesis