ChatPaper.aiChatPaper

SAM2Point: Segmentare qualsiasi 3D come video in modalità zero-shot e promptabile

SAM2Point: Segment Any 3D as Videos in Zero-shot and Promptable Manners

August 29, 2024
Autori: Ziyu Guo, Renrui Zhang, Xiangyang Zhu, Chengzhuo Tong, Peng Gao, Chunyuan Li, Pheng-Ann Heng
cs.AI

Abstract

Introduciamo SAM2Point, un'esplorazione preliminare che adatta il Segment Anything Model 2 (SAM 2) per la segmentazione 3D zero-shot e promptable. SAM2Point interpreta qualsiasi dato 3D come una serie di video multidirezionali e sfrutta SAM 2 per la segmentazione nello spazio 3D, senza ulteriore addestramento o proiezione 2D-3D. Il nostro framework supporta vari tipi di prompt, inclusi punti 3D, box e maschere, e può generalizzare attraverso scenari diversificati, come oggetti 3D, scene indoor, ambienti outdoor e LiDAR sparso grezzo. Dimostrazioni su molteplici dataset 3D, ad esempio Objaverse, S3DIS, ScanNet, Semantic3D e KITTI, evidenziano le robuste capacità di generalizzazione di SAM2Point. A nostra conoscenza, presentiamo l'implementazione più fedele di SAM in 3D, che potrebbe servire come punto di partenza per future ricerche sulla segmentazione 3D promptable. Demo online: https://huggingface.co/spaces/ZiyuG/SAM2Point . Codice: https://github.com/ZiyuGuo99/SAM2Point .
English
We introduce SAM2Point, a preliminary exploration adapting Segment Anything Model 2 (SAM 2) for zero-shot and promptable 3D segmentation. SAM2Point interprets any 3D data as a series of multi-directional videos, and leverages SAM 2 for 3D-space segmentation, without further training or 2D-3D projection. Our framework supports various prompt types, including 3D points, boxes, and masks, and can generalize across diverse scenarios, such as 3D objects, indoor scenes, outdoor environments, and raw sparse LiDAR. Demonstrations on multiple 3D datasets, e.g., Objaverse, S3DIS, ScanNet, Semantic3D, and KITTI, highlight the robust generalization capabilities of SAM2Point. To our best knowledge, we present the most faithful implementation of SAM in 3D, which may serve as a starting point for future research in promptable 3D segmentation. Online Demo: https://huggingface.co/spaces/ZiyuG/SAM2Point . Code: https://github.com/ZiyuGuo99/SAM2Point .
PDF282November 14, 2024