小艺

Tool introduction:

小艺It is the "Hongmeng Full Scenario AI Smart Assistant" independently developed by Huawei. The core relies on the Pangu Model and DeepSeek dual architecture to realize the evolution from a "passive tool" to an "active super assistant". The platform covers the web, mobile phones (Android/iOS) and Hongmeng full-scene devices (PC, smart screens, cars, etc.). It supports multi-modal interactions such as voice, vision, and circling. It has real-life dialogue, real-time environment recognition, and professional Learning assistance, cross-application task execution and other capabilities. As of August 2023, the number of monthly active users of 小艺 exceeds 200 million. After the upgrade in October 2024, the amount of knowledge exceeded one trillion, and the mission success rate exceeded 90%. At the same time, through end-side data processing (deleting immediately after calculation), the Institute of Information Technology AIGC privacy certification ensures the security of user data and is the core intelligent center of Hongmeng Ecology's "1+8+N" strategy.

Core functions:

• Real-life dialogue and dialect interaction: Support the recognition and response of 16 dialects (Shanghai dialect, southern Fujian, etc.), and the dialogue text is converted into dialects in real time; achieve emotional empathy based on a large psychological model, and have "interrupted and uninterrupted" full-duplex dialogue., as natural as chatting with old friends ​
• 小艺 See the world (Real-Time Communication Recognition): Call the camera to achieve Real-Time Communication interaction, support animal and plant/attraction identification, wearing suggestions, menu translation, and can also remind visually impaired users of obstacles, describe the environment, and assist travel ​
• In-depth problem solving and learning assistance: Covering Chinese, English, physics, chemistry and other subjects, supporting notebook selection and "drawing inferences from one example" difficulty selection; adding AI guidance writing to provide inspiration for composition and poetry creation to adapt to students 'learning scenarios ​
• Intelligent document processing: Support drag-and-drop upload of 50 types of files such as PDF and DOCX, extract key points with one click, generate mind maps, and translate foreign languages; connect local documents, you can extract information from the entire data and summarize answers, improving office efficiency. ​
• Cross-application task execution: Complete complex operations with the sentence "小艺 Help", such as "Jingdong repurchases cat food" and "navigates to Hongqiao Airport". Automatic cross-application linkage in the background. Only the payment link requires user confirmation; support customization Scenario automation (such as "Go to the company to open a small red book for silence") ​
• AI retouching and image optimization: Natural language instructions realize the elimination of passers-by, changing backgrounds, and repairing old photos, and support photoshop-level light effects such as surface light/side backlight; travel scenes automatically capture photos, locate and generate travel notes, and remind you to cache video when offline. ​
• 小艺 Huiji (meeting management): Transcribes meeting minutes in real time and generates summaries to support meeting hosting reminders; Hongmeng computer can create PPT in one sentence to improve office collaboration efficiency ​
• Hongmeng full-scene equipment control: Adjust the brightness of the computer screen, screen capture and recording in one sentence, or link smart screens and vehicle-machine circulation tasks (such as synchronizing mobile navigation to vehicle-machine) to achieve seamless collaboration across devices

Usage scenarios:

• Daily assistant scenarios: Check weather, book air tickets, check express delivery, and remember parking spaces. One-sentence instructions are quickly completed, supporting continuous conversations without repeated wake-ups ​
• Learning scenarios: Students use the in-depth problem-solving function to overcome subject problems, and use AI to guide writing to complete homework; when reading foreign documents, the "translation screen" realizes scrolling and mutual translation ​
• Office scenarios: Workplace people use documents to intelligently process summary reports and generate PPTs,小艺 Huiji records meeting key points, and cross-application tasks are executed to reduce manual operations ​
• Travel scenarios: Use "小艺 See the World" to identify attractions and translate menus while traveling, and remind you to cache videos before offline; start navigation with one sentence to automatically generate travel notes ​
• Life service scenarios: one-click re-purchase of commonly used products (such as cat food), customize equipment automation tasks (such as turning on air conditioning at home), voice control smart home ​
• Visually impaired auxiliary scene: "小艺 See the World" reminds online ride-hailing to arrive, identify obstacles, describe the park scenery, and facilitate barrier-free travel when going out.

Applicable population:

• Short Video creators/self-media people: Use AI to retouch and optimize materials, offline cache videos to ensure creation, and use imaging functions to improve content texture ​
• Ordinary consumers (Hongmeng device users): Daily information checking, life services, device control, enjoy a full-scene intelligent and convenient experience ​
• Student group: In-depth problem solving, AI writing assisted learning, foreign language translation improves reading efficiency, and meets the learning needs of primary and secondary schools to colleges and universities ​
• Workplace people: Document summary, meeting minutes, and cross-application task execution improve office efficiency and reduce cumbersome operations ​
• Visually impaired users: "小艺 See the World" assists travel, obstacle identification and environmental description solve travel pain points ​
• Travel enthusiasts: attraction identification, travel generation, offline video caching, fully adapted to the needs of travel scenarios

Unique advantages:

1. Hongmeng full-scene collaboration: Deeply integrated into Hongmeng OS, supporting task flow from mobile phones, PCs, vehicles and other devices (such as cross-device synchronization of navigation), achieving "one-place wake-up, multi-end response", far exceeding single-device AI assistants ​
2. Strict privacy protection: Based on the Hongmeng HPIC platform, data is processed on the end-side and deleted as soon as it is calculated, and data is isolated and calculated on the cloud; it has obtained the first batch of AIGC Privacy Excellence Certification from the Institute of Information and Technology, and the entire process ensures user information security ​
3. Deep empowerment of large models: Pangu Large Model +DeepSeek dual architecture supports, which can adaptively "think quickly"(such as checking the weather) or "think deeply"(such as complex reasoning) based on the complexity of the problem, and the answer is highly accurate. ​
4. Proactive service capabilities: Based on user habits to predict needs (such as setting an alarm clock after asking for the time late at night), customize scenarios to automate tasks, and upgrade from "passive response" to "active service" ​
5. Multimodal interaction is friendly: Covering voice, visual, circling, file drag-and-drop and other interactive methods, dialect support and emotional empathy reduce the threshold of use, making it easy for the elderly and visually impaired users to get started.