Zero-shot object detection and referring expression comprehension using vision-language models

This project focused on constructing a comprehensive perception pipeline integrating Natural Language Processing (NLP), zero-shot object detection, and Referring Expression Comprehension (ReC) within a ROS (Robot Operating System) framework. The aim was to enhance robotic assistive devices in accura...

Full description

Saved in:
Bibliographic Details
Main Author: A Manicka, Praveen
Other Authors: Ang Wei Tech
Format: Final Year Project
Language:English
Published: Nanyang Technological University 2024
Subjects:
Online Access:https://hdl.handle.net/10356/177827
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-177827
record_format dspace
spelling sg-ntu-dr.10356-1778272024-06-02T23:56:55Z Zero-shot object detection and referring expression comprehension using vision-language models A Manicka, Praveen Ang Wei Tech School of Mechanical and Aerospace Engineering Rehabilitation Research Institute of Singapore (RRIS) WTAng@ntu.edu.sg Computer and Information Science Engineering This project focused on constructing a comprehensive perception pipeline integrating Natural Language Processing (NLP), zero-shot object detection, and Referring Expression Comprehension (ReC) within a ROS (Robot Operating System) framework. The aim was to enhance robotic assistive devices in accurately interpreting natural language commands and grounding language to physical objects in the real world. To achieve this, we compared various combinations of zero-shot object detectors and ReC models, specifically specifically OWL-ViT and Grounding DINO for zero-shot object detection; and ReCLIP and GPT-4 for ReC. Our evaluation assessed the models' capabilities in counting, spatial reasoning, understanding superlatives, handling multiple instances, self-referential comprehension, and identifying household objects. The findings were showed that GPT-4 outperformed ReCLIP as for the purpose of ReC, and the combination of Grounding DINO and GPT-4 proved to be the best zero-shot object detector and ReC pair. Bachelor's degree 2024-05-31T12:13:12Z 2024-05-31T12:13:12Z 2024 Final Year Project (FYP) A Manicka, P. (2024). Zero-shot object detection and referring expression comprehension using vision-language models. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/177827 https://hdl.handle.net/10356/177827 en application/pdf Nanyang Technological University
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic Computer and Information Science
Engineering
spellingShingle Computer and Information Science
Engineering
A Manicka, Praveen
Zero-shot object detection and referring expression comprehension using vision-language models
description This project focused on constructing a comprehensive perception pipeline integrating Natural Language Processing (NLP), zero-shot object detection, and Referring Expression Comprehension (ReC) within a ROS (Robot Operating System) framework. The aim was to enhance robotic assistive devices in accurately interpreting natural language commands and grounding language to physical objects in the real world. To achieve this, we compared various combinations of zero-shot object detectors and ReC models, specifically specifically OWL-ViT and Grounding DINO for zero-shot object detection; and ReCLIP and GPT-4 for ReC. Our evaluation assessed the models' capabilities in counting, spatial reasoning, understanding superlatives, handling multiple instances, self-referential comprehension, and identifying household objects. The findings were showed that GPT-4 outperformed ReCLIP as for the purpose of ReC, and the combination of Grounding DINO and GPT-4 proved to be the best zero-shot object detector and ReC pair.
author2 Ang Wei Tech
author_facet Ang Wei Tech
A Manicka, Praveen
format Final Year Project
author A Manicka, Praveen
author_sort A Manicka, Praveen
title Zero-shot object detection and referring expression comprehension using vision-language models
title_short Zero-shot object detection and referring expression comprehension using vision-language models
title_full Zero-shot object detection and referring expression comprehension using vision-language models
title_fullStr Zero-shot object detection and referring expression comprehension using vision-language models
title_full_unstemmed Zero-shot object detection and referring expression comprehension using vision-language models
title_sort zero-shot object detection and referring expression comprehension using vision-language models
publisher Nanyang Technological University
publishDate 2024
url https://hdl.handle.net/10356/177827
_version_ 1800916371569115136