ProDyGS: Dynamic Gaussian Splatting From a Single Static Monocular Camera Hits SOTA
kwangmoo_yi · x · 2026-09-30
ProDyGS (Univ. of Bologna + Sony Depthsensing, IROS 2026) tackles dynamic novel view synthesis without multi-view supervision. It estimates temporally consistent monocular depth (Depth Pro + neural disparity refinement, scale-shift stabilization on static background), then uses 3D Gaussians to render proxy images from arbitrary viewpoints as pseudo-multi-view supervision, with a deformation network learning temporal dynamics. On DyNeRF it achieves SOTA, outperforming methods using stronger priors like scene flow (e.g., MoDGS). Code coming soon.
Related event: ProDyGS Enables Dynamic Gaussian Splatting from a Single Static Camera(2 posts)→
More from Multimodal
- Four imaginary tokens for Midjourney v8.2 produce memory ghosts and bone echoes — LudovicCreator · 2026-09-30
- Hyper-personalized music is BS: music is culture and inherently social, argues developer — jordiponsdotme · 2026-09-30
- NUS Proposes StoryEngine: A State-Grounded Agentic Framework for Coherent Long-Form Video Storytelling — NationalUniversityofSingapore · 2026-09-30
- One Year of Local Image Generation: Why Civitai and ComfyUI Both Fall Short — BenDLH · 2026-09-30
- Opus made a launch video for Violetto 1B in 50 minutes amid zero media coverage — tensorqt · 2026-09-30
- Meshy hits $100M ARR in under two years as GPT-6 Astra stirs the AI 3D debate — 量子位 · 2026-09-30