Radu Beche, Raoul de Charette · 2026 · arXiv (Cornell University)
DOI: 10.48550/arxiv.2610.28615 · cited by 0 works (OpenAlex) · open-access PDF
في هذا المستوى المستمر من التفاصيل (LoD) الإعدادية، التحدي هو تحديد ما Gaussians يجب الحفاظ على القدرة عندما يقلل من ميزانية الحوسبة، دراسة تجربة فريدة حول 3DGS-MCMC التكوينات تشير إلى أن صعوبة الصورة، التي تم تحديدها من وجهة نظر الدخول قبل التحسين، تحدد حيث يتم تركيب خطأ التكوين في الصورة والمنطقة 3D. كما يظهر أيضا أن المناطق البسيطة تحتوي على الكثير من Gaussians وتخفيض كمية كبيرة من Gaussians في ميزانية أقل، على الرغم من أنها أقل تكلفة من حيث الجانب. بعد ذلك، في KoalaGS، أثناء التدريب، نحن نستخدم هذه الصعوبة المهمة لتجنب التخفيضات في مناطق التكوينات
Machine translation (Meta M2M100); the English original follows.
Deploying a trained 3D Gaussian Splatting scene under shrinking compute budgets requires one primitive set to remain effective when rendered at smaller prefixes. In this continuous level of detail (LoD) setting, the challenge is to decide which Gaussians should retain capacity as the budget decreases. A pilot study on 3DGS-MCMC reconstructions shows that image complexity, computed from the input views before optimization, predicts where reconstruction error concentrates in image and 3D space. It also reveals that simple regions contain too many Gaussians and degrade as much as complex ones at lower budgets, despite being inherently cheaper to represent. Subsequently, in KoalaGS, during training we use this complexity prior to bias relocation toward poorly reconstructed areas while ensuring smooth degradation across LoDs. A primitive importance score then enables inference at low-capacity, while preserving the scene structure. Tested across three datasets, KoalaGS maintains competitive full-capacity quality while significantly improving quality across the budget spectrum, yielding smoother transitions across LoDs, and enabling faster progressive streaming.
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
Topics: Computer science, Artificial intelligence, Gaussian, Inference, Set (abstract data type), Computer vision
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.28615, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange