AI-Generated Video Practices for International Hotels
At the WOT2024 conference, Qunar Travel’s CTO Zheng Jimin presented a comprehensive overview of AI-generated video production for international hotels, detailing challenges, AI-driven workflow automation, practical implementation steps, multilingual translation enhancements, and performance results, offering valuable insights for scaling high‑quality hotel video content.
This article compiles highlights from Zheng Jimin, CTO of Qunar Travel, who delivered a keynote titled “AI-Generated Video Practices for International Hotels” at the WOT2024 Global Technology Innovation Conference.
Video Generation Challenges and Opportunities
The team identified low video coverage (≈19.6%) for international hotels and recognized the potential of AIGC to generate videos that improve conversion rates. Key challenges include selecting high‑quality images, handling multilingual user reviews, and ensuring diverse and accurate content.
Professional Film Production Process AI‑ification
The traditional four‑step workflow—planning, storyboard creation, shooting, and post‑production—is outlined, with each step illustrated by diagrams. AI is applied to automate storyboard generation, image and text preprocessing, and seamless video stitching.
AI‑Generated Video Practice
Quality assessment focuses on value, visual clarity (1080p/4K), and thematic relevance. The production pipeline consists of four steps: material selection, preprocessing, storyboard creation, and template‑based editing. Multilingual translation leverages large language models (GPT‑3.5/4) to improve fluency and style across 27 languages.
Multimodal Generation Enhancements
The team experimented with Pika and Runway platforms, finding Runway’s Gen‑2 model most effective for realistic sea‑wave dynamics while emphasizing the need for physical‑logic constraints. Parameter tuning remains critical despite powerful tools.
Video Generation Results and Reflections
Examples showcase various styles—minimalist business, island, and Japanese‑style hotels—demonstrating customized templates, dynamic effects, and localized highlights (e.g., infinity pools). Deployment in the app increased video play‑through rates by about 6%.
Summary
The project learned that 1080p balances quality and mobile load time, background music often outperforms narrated voice‑overs for high‑end hotels, and physical realism in dynamic images is essential. Future plans include fully customized video content per hotel style, targeting different guest segments, with a production cost of about ¥1.25 per video and a turnaround of 30‑60 seconds.
Qunar Tech Salon
Qunar Tech Salon is a learning and exchange platform for Qunar engineers and industry peers. We share cutting-edge technology trends and topics, providing a free platform for mid-to-senior technical professionals to exchange and learn.
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.