Kestrel: 3D Multimodal LLM for Part-Aware Grounded Description

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ahmed, Mahmoud, Fei, Junjie, Ding, Jian, Bakr, Eslam Mohamed, Elhoseiny, Mohamed
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913973639577600
author Ahmed, Mahmoud
Fei, Junjie
Ding, Jian
Bakr, Eslam Mohamed
Elhoseiny, Mohamed
author_facet Ahmed, Mahmoud
Fei, Junjie
Ding, Jian
Bakr, Eslam Mohamed
Elhoseiny, Mohamed
contents In this paper, we introduce Part-Aware Point Grounded Description (PaPGD), a challenging task aimed at advancing 3D multimodal learning for fine-grained, part-aware segmentation grounding and detailed explanation of 3D objects. Existing 3D datasets largely focus on either vision-only part segmentation or vision-language scene segmentation, lacking the fine-grained multimodal segmentation needed for robotic navigation and interaction in real-world environments. To address this gap, we present the 3DCoMPaT Grounded Instructions (3DCoMPaT-GrIn) Dataset, a comprehensive resource that pairs rich point cloud descriptions with corresponding part-level segmentation masks. This dataset encompasses extensive samples designed for both PaPGD and fine-grained single-part grounding tasks. To tackle the inherent challenges of grounding objects and generating grounded descriptions at the part level, we propose Kestrel, a part-aware 3D multimodal large language model that integrates an advanced language model for nuanced language comprehension with multi-level point feature propagation and query refinement mechanism to enhance spatial reasoning at the part level. The extensive experiments demonstrate that Kestrel effectively bridges the gap between part-aware language understanding and 3D segmentation grounding, paving the way for more robust and interpretable 3D object comprehension that meets the demands of real-world robotic applications. Project page at https://feielysia.github.io/Kestrel.github.io/
format Preprint
id arxiv_https___arxiv_org_abs_2405_18937
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Kestrel: 3D Multimodal LLM for Part-Aware Grounded Description
Ahmed, Mahmoud
Fei, Junjie
Ding, Jian
Bakr, Eslam Mohamed
Elhoseiny, Mohamed
Computer Vision and Pattern Recognition
Computation and Language
In this paper, we introduce Part-Aware Point Grounded Description (PaPGD), a challenging task aimed at advancing 3D multimodal learning for fine-grained, part-aware segmentation grounding and detailed explanation of 3D objects. Existing 3D datasets largely focus on either vision-only part segmentation or vision-language scene segmentation, lacking the fine-grained multimodal segmentation needed for robotic navigation and interaction in real-world environments. To address this gap, we present the 3DCoMPaT Grounded Instructions (3DCoMPaT-GrIn) Dataset, a comprehensive resource that pairs rich point cloud descriptions with corresponding part-level segmentation masks. This dataset encompasses extensive samples designed for both PaPGD and fine-grained single-part grounding tasks. To tackle the inherent challenges of grounding objects and generating grounded descriptions at the part level, we propose Kestrel, a part-aware 3D multimodal large language model that integrates an advanced language model for nuanced language comprehension with multi-level point feature propagation and query refinement mechanism to enhance spatial reasoning at the part level. The extensive experiments demonstrate that Kestrel effectively bridges the gap between part-aware language understanding and 3D segmentation grounding, paving the way for more robust and interpretable 3D object comprehension that meets the demands of real-world robotic applications. Project page at https://feielysia.github.io/Kestrel.github.io/
title Kestrel: 3D Multimodal LLM for Part-Aware Grounded Description
topic Computer Vision and Pattern Recognition
Computation and Language
url https://arxiv.org/abs/2405.18937