Functionality understanding and segmentation in 3D scenes

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Corsetti, Jaime, Giuliari, Francesco, Fasoli, Alice, Boscaini, Davide, Poiesi, Fabio
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918036959657984
author Corsetti, Jaime
Giuliari, Francesco
Fasoli, Alice
Boscaini, Davide
Poiesi, Fabio
author_facet Corsetti, Jaime
Giuliari, Francesco
Fasoli, Alice
Boscaini, Davide
Poiesi, Fabio
contents Understanding functionalities in 3D scenes involves interpreting natural language descriptions to locate functional interactive objects, such as handles and buttons, in a 3D environment. Functionality understanding is highly challenging, as it requires both world knowledge to interpret language and spatial perception to identify fine-grained objects. For example, given a task like 'turn on the ceiling light', an embodied AI agent must infer that it needs to locate the light switch, even though the switch is not explicitly mentioned in the task description. To date, no dedicated methods have been developed for this problem. In this paper, we introduce Fun3DU, the first approach designed for functionality understanding in 3D scenes. Fun3DU uses a language model to parse the task description through Chain-of-Thought reasoning in order to identify the object of interest. The identified object is segmented across multiple views of the captured scene by using a vision and language model. The segmentation results from each view are lifted in 3D and aggregated into the point cloud using geometric information. Fun3DU is training-free, relying entirely on pre-trained models. We evaluate Fun3DU on SceneFun3D, the most recent and only dataset to benchmark this task, which comprises over 3000 task descriptions on 230 scenes. Our method significantly outperforms state-of-the-art open-vocabulary 3D segmentation approaches. Project page: https://tev-fbk.github.io/fun3du/
format Preprint
id arxiv_https___arxiv_org_abs_2411_16310
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Functionality understanding and segmentation in 3D scenes
Corsetti, Jaime
Giuliari, Francesco
Fasoli, Alice
Boscaini, Davide
Poiesi, Fabio
Computer Vision and Pattern Recognition
Understanding functionalities in 3D scenes involves interpreting natural language descriptions to locate functional interactive objects, such as handles and buttons, in a 3D environment. Functionality understanding is highly challenging, as it requires both world knowledge to interpret language and spatial perception to identify fine-grained objects. For example, given a task like 'turn on the ceiling light', an embodied AI agent must infer that it needs to locate the light switch, even though the switch is not explicitly mentioned in the task description. To date, no dedicated methods have been developed for this problem. In this paper, we introduce Fun3DU, the first approach designed for functionality understanding in 3D scenes. Fun3DU uses a language model to parse the task description through Chain-of-Thought reasoning in order to identify the object of interest. The identified object is segmented across multiple views of the captured scene by using a vision and language model. The segmentation results from each view are lifted in 3D and aggregated into the point cloud using geometric information. Fun3DU is training-free, relying entirely on pre-trained models. We evaluate Fun3DU on SceneFun3D, the most recent and only dataset to benchmark this task, which comprises over 3000 task descriptions on 230 scenes. Our method significantly outperforms state-of-the-art open-vocabulary 3D segmentation approaches. Project page: https://tev-fbk.github.io/fun3du/
title Functionality understanding and segmentation in 3D scenes
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2411.16310