Iris-Z 如何以生成模型超越專用幾何模型
Iris-Z: Scaling One Generative Model Past Specialist Geometry Models
Iris-Z 是一個同時生成圖像外觀與深度的通用圖像生成模型,以單一骨幹網絡處理圖像生成、深度估算及依據深度繪製圖像。在足夠規模下,它在既有深度基準測試中超越專用模型,同時保留原有骨幹網絡的圖像生成能力;研究團隊向社羣發布了兩個模型。
Published on Nov 11
Authors:
,
Abstract
A unified generative backbone jointly models image appearance and depth, surpassing specialized depth models at scale while preserving general image generation.
Language moved from many task-specific models to one unified model: a single generative model with a single generative training task that handles all of them within the same backbone, and keeps improving with data, parameters and compute. We believe the bitter lesson applies just as much to computer vision, 3D and spatial understanding, in the end to visual intelligence: the effort should go into general architectures that solve all these tasks in a unified way, and into scaling them, rather than into task-specific designs. We test this on monocular geometry, a task that has belonged to specialised architectures for years. Iris-Z is a general image generator that produces appearance and geometry together. A single backbone denoises, reasons about and understands the image and its depth at the same time, and one set of weights generates images with their depth, estimates depth from photographs and paints images onto a given depth. At enough scale it surpasses state-of-the-art specialists on established depth benchmarks while keeping the generative abilities of the backbone it started from without specialized layers. We release two models for the community.
View arXiv page View PDF Project page Add to collection
Community
Paper submitter
Models citing this paper 3
Depth Estimation • Updated 3 days ago • 40speridlabs/iris-z
Text-to-Image • Updated 1 day ago • 29eduardoabreu81/iris-z-small-quantized
Text-to-Image • Updated 1 day ago • 21eduardoabreu81/iris-z-small-turbo
Datasets citing this paper 0
No dataset linking this paper
Cite arxiv.org/abs/2611.16010 in a dataset README.md to link it from this page.
Spaces citing this paper 1
Collections including this paper 0
No Collection including this paper
Add this paper to a collection to link it from this page.
來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co