Enterprise AI Analysis
Chang-E: A High-Quality Motion Capture Dataset of Chinese Classical Dunhuang Dance
Derived from the mural drawings in the UNESCO-listed Mogao Caves, Dunhuang dance has unique cultural value but faces challenges of digitization and preservation. In this article, we introduce the first open comprehensive motion capture dataset of Dunhuang dance, Chang-E, including full-body movements documented across 8 categories, totaling 40 minutes of professional dance (preview available at https://cislab.hkust-gz.edu.cn/projects/chang-e/). This dataset contains three formats: skeleton data acquired from motion capture, body mesh generated from skeleton using machine learning, and multiview videos recorded on site. The dataset supports various creative applications for Dunhuang dance culture, as demonstrated by an immersive new media exhibition. Through the curation process, we applied motion inbetweening algorithms to concatenate different dance sequences for choreography. Also, these reinterpreted dance sequences are synchronized with music using retiming techniques, augmenting the rhythms and harmony between the music and dance performance. Furthermore, we applied visual effects on the regenerated motion sequences of digital dancers, achieving artistic and appealing visual results echoing Buddhist discourses of meditation and bodily cognition. The Chang-E dataset enables digital preservation and creative reimagination of Dunhuang dance, offering not only high-quality data but also an interdisciplinary collaboration framework for future graphics and cultural heritage research.
Executive Impact Summary
The Chang-E dataset addresses a critical need for digitizing and preserving Dunhuang dance, a unique cultural heritage facing preservation challenges. By leveraging advanced motion capture technology, machine learning, and multi-view video, this project delivers a comprehensive, high-quality dataset that captures the full-body movements of professional dancers across 8 distinct categories, totaling 40 minutes of recorded dance. This dataset not only serves as a valuable resource for digital preservation but also enables diverse creative applications, including immersive new media exhibitions, motion inbetweening for choreography, and synchronized dance-music performances. The use of SMPL models for mesh generation ensures realistic animated human shapes and natural deformations, laying a robust foundation for future research in computer graphics, cultural heritage, and AI-driven generative dance forms. The project’s interdisciplinary approach, culminating in a successful exhibition, highlights the potential of combining traditional art with cutting-edge digital technology for broader public engagement and cultural dissemination.
Minutes of High-Quality MoCap Data
Dance Categories Documented
Data Formats (Skeleton, Mesh, Video)
Frames Optimized per 10 min (RTX3090 GPU)
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
Motion Capture & Data Processing
This section details the methodologies for capturing and processing motion data, including the use of passive optical motion capture systems, marker placement strategies, and the conversion of BVH data to SMPL format. It highlights the challenges of accurately capturing complex dance movements and the post-processing techniques applied to ensure data quality and realism.
Dunhuang Dance Heritage
Explores the unique cultural value of Dunhuang dance, its origins from Mogao Caves murals, and the challenges in its digital preservation. It emphasizes the historical significance and aesthetic elements of Dunhuang dance, contrasting it with other Chinese dance forms and highlighting the need for dedicated datasets.
Creative Applications & Exhibitions
Focuses on the practical applications of the Chang-E dataset, such as generating dance sequences for immersive exhibitions, motion inbetweening for choreography, and synchronizing dance with music. It also covers the integration of visual effects to enhance digital performances and the successful public exhibition at Harvard University and Shanghai.
6890
SMPL Model Vertices for High-Fidelity Mesh
Enterprise Process Flow
Mural Image/Stage Drama
→
Motion Capture
→
BVH Data
→
SMPL Parameters Fitting
→
Chang-E Dataset Output
| Feature | Chang-E Dataset | General MoCap Datasets (e.g., Human3.6M, AMASS) | NUS Chinese Dance Dataset |
|---|---|---|---|
| Dance Genre | Chinese Classical Dunhuang Dance | Diverse human activities, street dances | Limited Chinese dance segments |
| Cultural Context | Deeply rooted in Buddhist art & Mogao Caves | General human motion; less cultural specificity | Generic traditional Chinese dance |
| Data Quality & Detail | High-quality, professional dance capture, full-body mesh & skeleton | High-quality for common actions; diverse range | Short segments, less depth for specific forms |
| Dataset Focus | Dedicated to Dunhuang dance (8 categories, 40 min) | Broad spectrum of motion types; large scale | Limited segments (7 segments, <30s each) |
| Applications | Digital preservation, creative reimagination, immersive exhibitions, AI generative dance | Character animation, digital humans, machine learning, robotics | Basic dance research, limited creative apps |
Immersive 'Cave Dance' Exhibition Success
The 'Cave Dance' exhibition, hosted at Harvard University and Shanghai, showcased the Chang-E dataset's transformative potential. Leveraging motion inbetweening, dance-music synchronization, and advanced visual effects, the exhibition created an immersive environment where visitors experienced Dunhuang dance in a digital, dynamic, and lifelike reimagination. The exhibition garnered significant attention, demonstrating how modern technology can rejuvenate ancient art forms and promote cultural heritage.
This initiative highlights a successful interdisciplinary collaboration between dance and computing communities. It provides a blueprint for future cultural heritage projects, proving that high-quality motion capture data combined with innovative digital applications can significantly enhance public engagement and ensure the longevity of intangible cultural assets.
Advanced ROI Calculator
Estimate your enterprise's potential annual savings and reclaimed human hours with AI automation.
Estimated Annual Savings
$0
Annual Hours Reclaimed
0
Phased AI Implementation Roadmap
A structured approach to integrating AI, from data acquisition to full deployment and cultural dissemination.
Phase 1: Data Acquisition & Pre-processing
High-speed motion capture of 8 Dunhuang dance categories by professional dancers, followed by initial data cleaning, marker tracking, and reconstruction of 3D feature points.
Phase 2: BVH to SMPL Conversion & Grounding
Conversion of raw BVH skeleton data to the SMPL model format, including fitting 3D key points to SMPL parameters and implementing a grounding algorithm to ensure realistic foot-ground interaction.
Phase 3: Creative Applications Development
Development of core applications such as motion inbetweening for seamless choreography, dance-music synchronization using retiming techniques, and integration of various visual effects (particles, Bodhisattva skin, stick figures) to enhance digital performances.
Phase 4: Exhibition & Dissemination
Deployment of the immersive 'Cave Dance' exhibition, showcasing the reimagined Dunhuang dance sequences to the public at prestigious venues, fostering interdisciplinary collaboration, and promoting cultural preservation.
Ready to Transform Your Enterprise with AI?
Leverage cutting-edge AI insights to drive efficiency, innovation, and strategic growth. Book a personalized consultation with our experts today.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat