Zhenjun Zhao (@zhenjun_zhao)
2025-06-24 | ❤️ 38 | 🔁 2
BulletGen: Improving 4D Reconstruction with Bullet-Time Generation
@DRozumnyi, @JonathonLuiten, Numair Khan, Johannes Schönberger, Peter Kontschieder
tl;dr: generative video diffusion model enhances 4D Gaussian-based reconstruction at a single frozen “bullet-time” step
https://arxiv.org/abs/2506.18601
🔗 Related
See similar notes in domain-vision-3d, domain-genai