Media
Featured at CMG's "AI Mid-Autumn" media arts event
Our Into-the-Painting experience was showcased at CMG's "AI Mid-Autumn" conference (Jing'an Joy City, Shanghai), with the host trying it live on air.
Self-developed generation algorithms · Enterprise delivery
Self-Developed Video & Image Generation Models — Enterprise Licensing & Deployment
Self-developed video and image generation algorithms, delivering integrable, batchable, and reviewable generation capabilities for enterprises.
First/last-frame controlMulti-subject referenceBatch generationPrivate deployment
Proven results
50k+
Users served
2
National 5A scenic areas
3
Representative products
2
Silicon Valley partners
Algorithm → License → Delivery
00 · Newsroom
Company news
Product launches, industry deployments, and partnerships — all in one place.
Launch
"Into the Painting" immersive device officially released
Built on our in-house visual generation algorithms, the Into-the-Painting device is now available for scenic areas, malls, and exhibitions — low-cost to deploy, placing visitors inside classic paintings instantly.
Deployment
AI portrait system live at National 5A scenic areas
The smart photography system is running at Yuelu Mountain and Orange Isle (both National 5A scenic areas) and Yangtze River Night City, with 50,000+ visitors served.
Recognition
Won a CCF Class-A startup competition award
Third prize in the 2024 Ma'anshan "AI + Industrial Manufacturing" track — our in-house R&D recognized by a leading authority.
Partnership
Long-term partnerships with SocialBook and AKOOL
We keep delivering video and image foundation model R&D for Silicon Valley AI companies, bringing our core algorithms to the global market.
Press & interviews
01 · Smart hardware · New
Into the Painting: live inside a classic scroll
An immersive interactive experience built on AI visual generation — visitors step in front of the device and are instantly woven into a classic painting. Low-cost to deploy for scenic areas, museums, and cultural projects.
What is "Into the Painting"?
Stand in front of the device and choose your favorite classical painting scene. The system recognizes your facial features and combines AI image generation, expression driving, and portrait blending to place you inside the artwork instantly — you are in the painting, and the painting moves with you.
Technology
4K ultra-HD capture combined with AI face recognition, portrait segmentation, image generation, style transfer, expression driving, and real-time rendering — blending people naturally into the painting style and bringing static artworks to life.
Why it matters
Beyond the traditional photo op, it fuses cultural storytelling with AI: visitors walk away with a shareable, culture-rich digital keepsake, boosting engagement, interactivity, and brand exposure for scenic areas, museums, and cultural projects.
- Low-cost deployment
- Instant blending
- On-site keepsake
Phone lines are open on business days 9:00–18:00, or leave a message via the form below and we will get back to you shortly.
How it works
Pick a scene
Choose from classic painting scenes in front of the device
AI blending
Face recognition, image generation, and expression driving bring the artwork to life with you in it
Scan & take away
Generate a culture-rich digital keepsake and save or share it with a scan
Where it fits
Scenic areasMuseums & exhibitsMall pop-upsTrade showsBrand campaigns
02 · Algorithm Capabilities
Full-chain generation algorithms × Director Agent
From text-to-video to one-click Director Agent films — every capability defines clear inputs and deliverables, usable standalone or combined into a complete content pipeline.
Text-to-Video Algorithm
Generate coherent video clips from text prompts with style and camera control, multiple variants, and configurable first/last-frame and subject references for ads, creative exploration, and content production.
- Prompt semantic understanding
- Style & camera control
- Multiple variants
PROMPT 01
A 6-second product video with a night city background
PROMPT 02
Three static product images adapted for ad placements
PROMPT 03
Multiple camera and pacing variants
VID-02
Image-to-Video Algorithm
Animate product images, character images, and scene references while preserving subject and temporal consistency, with camera motion and multi-subject reference support.
- Subject & temporal consistency
- Camera motion
- Multi-subject reference
IMG-03
Image Generation Algorithm
Text-to-image, image-to-image, and image editing with multi-size adaptation and style preservation, covering the full pipeline from static visuals to dynamic content.
- Text-to-image
- Image-to-image & editing
- Multi-size output
DIR-04
Director Agent System
A complete Director Agent handles the whole journey from idea to finished film: scriptwriting, storyboard design, shot scheduling, scoring, and editing — cinematic results in one click.
- AI script & storyboard
- Shot planning, scoring & editing
- One-click production
BAT-05
Production-scale content
From single samples to production-scale output, with batch jobs, priorities, and elastic compute — costs stay predictable.
- Batch job submission
- Priority & concurrency control
- Elastic compute
TUN-06
Fine-tuning & style customization
Directional fine-tuning on your brand assets and style samples keeps results on-brand, with a dedicated style library.
- Brand-style fine-tuning
- Dedicated style templates
- Continuous tuning
03 · Use Cases
Delivered per industry, from sample validation to production
Each use case maps to clear inputs and deliverables: validate with real samples first, then license or deploy.
FR-01
Ecommerce Content
Turn product images and selling points into short videos and display images for product pages, ads, and marketplace launches.
Product images in → videos & multi-size images out, raising SKU content coverage
FR-02
Advertising Creatives
Generate multiple shots, pacing, and composition variants around one campaign idea for faster A/B testing and asset pipelines.
Creative brief in → ad variants out, shortening creative cycles
FR-03
Animation & Short-Form Previews
Turn storyboards and character references into shot previews and atmosphere clips for planning and production communication.
Storyboards in → preview clips out, lowering iteration cost
FR-04
Game & Virtual Assets
Generate dynamic showcases for characters, props, and scenes for promotion, review, and asset management.
Static assets in → motion showcases out, improving asset usability
FR-05
Education & Training
Convert course scripts, knowledge points, and illustrations into explainer videos and teaching images for scalable production.
Course material in → videos & images out, cutting production cost
FR-06
Embedded in Your Platform
Embed generation capabilities into content platforms, SaaS, and automation tools, so end users create inside your product.
Your product in → creation capability out, generation becomes a selling point
04 · Delivery Flow
From sample validation to production, in five clear steps
Every step has a clear deliverable — move forward only when the results meet the bar.
TC-01
Requirements review
We align on what to generate, input assets, output specs, acceptance criteria, and timeline.
Requirement list & feasibility assessment
TC-02
Sample validation
We prove the generation quality on your real assets, with fine-tuning if needed; both sides confirm the bar.
Samples & acceptance criteria
TC-03
Agreement & licensing
Choose the cooperation model — standard licensing, private deployment, or custom R&D — and define scope and term.
Agreement & licensing plan
TC-04
Deployment & integration
We connect generation into your business systems and production workflow, with joint testing and a trial run.
Production go-live
TC-05
Production & optimization
Batch content production starts; quality and performance keep iterating with smooth version upgrades.
Delivery package & upgrade service
05 · Industry Deployment
From algorithm R&D to industry deployment: three representative AI products
Our AIGC technology is live in tourism, civic services, and international markets. All products below are developed and iterated in-house.
50K+
cumulative experiences
3
representative AI products
2
national 5A scenic areas
2
long-term Silicon Valley partners
SHOT-01
AI Ancient-Style Photo Booth System
An AI-powered ancient-style portrait system for tourism, deployed at Yuelu Mountain and Orange Isle national 5A scenic areas in Changsha, plus the Yangtze Night City in Ma'anshan, with over 50,000 cumulative experiences and purchases.
Tourism · 50,000+ cumulative experiences
SHOT-02
AI Life-Review Digital Tribute System
For digital remembrance, this system generates a high-quality life-review video automatically from a single photo and the person's biography. It has entered procurement programs at funeral homes in Changsha and Loudi.
Civic services · innovative digital tribute
SHOT-03
International Foundation Model R&D & Customization
We provide video and image foundation model R&D and customization for overseas customers, with long-term cooperation agreements signed with Silicon Valley AI companies SocialBook and AKOOL.
International · core algorithms exported
06 · Enterprise-grade service
From sample validation to production scale, fully covered
Verifiable results, dependable stability, and continuous iteration — lowering your integration and production risks.
CH-01
Low-friction start
No in-house algorithm team required. We provide a complete onboarding plan and dedicated technical support to get your first generation pipeline running fast.
CH-02
Proof first, acceptance-driven
We validate generation quality on your real samples first. Production starts only after both sides agree on acceptance criteria — what you see is what you get.
CH-03
Production at scale
From single samples to large-scale content output, with batch generation and compute scaling — costs stay predictable as volume grows.
CH-04
Data security & compliance
Private deployment keeps your data inside your environment, meeting data sovereignty and compliance audit requirements.
CH-05
Reliable operations
Generation services are monitored end to end, with automatic handling and retries for failed tasks, supporting enterprise-grade stability.
CH-06
Continuous improvement
Algorithms keep iterating — quality and performance improve steadily, with smooth, transparent version upgrades.
Proven deployments
07 · About HuiJie Tech
HuiJie Tech: In-house AIGC foundation models, deployed in real industries
Focused on generative AI foundation models and industry deployment
Founded in 2025, HuiJie Tech is a technology company focused on AIGC foundation model R&D, inference acceleration, and industry application. We develop video and image foundation models in-house — real-time inference, high-performance deployment, and multimodal content generation — following a model of core technology R&D, industry deployment, and international collaboration.
How We Develop
We build technology in-house and turn it into deployed products across tourism, civic services, and international markets, continuously pushing through bottlenecks in real-time deployment, compute optimization, and industry adoption.
CCF Class-A Entrepreneurship Award
3rd prize, 2024 Ma'anshan 'AI + Industrial Manufacturing' special track
Standard licensing
Private Deployment
Custom R&D
Deliver through real deployed cases, not conceptual promises
Focus on real-time inference and high-performance deployment for video & image foundation models
Balance generation quality, cost, latency, and stability
Self-developed technology, with international technology export
08 · FAQ
FAQ
Questions about algorithm capabilities, delivery models, licensing, and custom R&D
We are an in-house algorithm team. Founded in 2025, HuiJie Tech focuses on AIGC foundation model R&D, inference acceleration, and industry deployment. Our text-to-video, image-to-video, and image generation algorithms are developed and iterated internally, covering the full path from model research to industry delivery.
Deliver our in-house generation algorithms in the model you need
Tell us your scenario: what to generate, what inputs you have, and where results enter your workflow. We will propose a POC path and licensing plan.
Request a proposal
Helpful inputs: sample images, prompts, output specs, and concurrency estimates
09 · Contact
Contact Us
Tell us what video or image you want to generate, what inputs you have, and how you want to integrate. We will propose a sample validation and licensing plan.
Email
Service Coverage
Video & image generation algorithm licensing, private deployment, and custom R&D