SIGGRAPH Spotlight 90 – Kavita Bala and Hui Huang on Influential Computer Graphics Research
Jul 29, 2025
Kavita Bala, a Professor of Computer Science and Provost at Cornell University, and Hui Huang, a researcher from Sun Yat-sen University, dive into their influential work in computer graphics. They discuss the journey from their groundbreaking research to successful applications and startups. The conversation highlights advancements in fine-grained visual recognition, the evolution of AI integration in 3D graphics, and innovative data representation techniques. Their insights are a treasure trove for aspiring researchers seeking to make an impact in this dynamic field.
AI Snips
Chapters
Transcript
Episode notes
Hui Huang's Personal Milestone
- Hui Huang shared a personal story of submitting his SIGGRAPH paper just days after his daughter's birth.
- This moment linked a significant life event with a professional milestone, making the award especially meaningful.
Early Deep Learning Innovation
- Kavita Bala's paper was ahead of its time due to early deep learning use and contrastive loss formulation.
- Fine-grained recognition remains a compelling, unsolved problem important for AR/VR and beyond.
Robust Skeletons from Raw Data
- Hui Huang's method used the robust L1 median locally on raw point clouds without surface reconstruction.
- This made skeleton extraction practical for noisy, incomplete real-world 3D data unlike prior mesh-dependent methods.