Robotic-CLIP: Fine-tuning CLIP on Action Data for Robotic Applications

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Nguyen, Nghia, Vu, Minh Nhat, Ta, Tung D., Huang, Baoru, Vo, Thieu, Le, Ngan, Nguyen, Anh
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866914957939965952
author Nguyen, Nghia
Vu, Minh Nhat
Ta, Tung D.
Huang, Baoru
Vo, Thieu
Le, Ngan
Nguyen, Anh
author_facet Nguyen, Nghia
Vu, Minh Nhat
Ta, Tung D.
Huang, Baoru
Vo, Thieu
Le, Ngan
Nguyen, Anh
contents Vision language models have played a key role in extracting meaningful features for various robotic applications. Among these, Contrastive Language-Image Pretraining (CLIP) is widely used in robotic tasks that require both vision and natural language understanding. However, CLIP was trained solely on static images paired with text prompts and has not yet been fully adapted for robotic tasks involving dynamic actions. In this paper, we introduce Robotic-CLIP to enhance robotic perception capabilities. We first gather and label large-scale action data, and then build our Robotic-CLIP by fine-tuning CLIP on 309,433 videos (~7.4 million frames) of action data using contrastive learning. By leveraging action data, Robotic-CLIP inherits CLIP's strong image performance while gaining the ability to understand actions in robotic contexts. Intensive experiments show that our Robotic-CLIP outperforms other CLIP-based models across various language-driven robotic tasks. Additionally, we demonstrate the practical effectiveness of Robotic-CLIP in real-world grasping applications.
format Preprint
id arxiv_https___arxiv_org_abs_2409_17727
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Robotic-CLIP: Fine-tuning CLIP on Action Data for Robotic Applications
Nguyen, Nghia
Vu, Minh Nhat
Ta, Tung D.
Huang, Baoru
Vo, Thieu
Le, Ngan
Nguyen, Anh
Robotics
Computer Vision and Pattern Recognition
Vision language models have played a key role in extracting meaningful features for various robotic applications. Among these, Contrastive Language-Image Pretraining (CLIP) is widely used in robotic tasks that require both vision and natural language understanding. However, CLIP was trained solely on static images paired with text prompts and has not yet been fully adapted for robotic tasks involving dynamic actions. In this paper, we introduce Robotic-CLIP to enhance robotic perception capabilities. We first gather and label large-scale action data, and then build our Robotic-CLIP by fine-tuning CLIP on 309,433 videos (~7.4 million frames) of action data using contrastive learning. By leveraging action data, Robotic-CLIP inherits CLIP's strong image performance while gaining the ability to understand actions in robotic contexts. Intensive experiments show that our Robotic-CLIP outperforms other CLIP-based models across various language-driven robotic tasks. Additionally, we demonstrate the practical effectiveness of Robotic-CLIP in real-world grasping applications.
title Robotic-CLIP: Fine-tuning CLIP on Action Data for Robotic Applications
topic Robotics
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2409.17727