Skip to main navigation Skip to search Skip to main content

GRiT: A Generative Region-to-Text Transformer for Object Understanding

  • Jialian Wu
  • , Jianfeng Wang
  • , Zhengyuan Yang
  • , Zhe Gan
  • , Zicheng Liu
  • , Junsong Yuan
  • , Lijuan Wang
  • SUNY Buffalo
  • Advanced Micro Devices
  • Microsoft USA
  • Apple

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

16 Scopus citations

Abstract

This paper presents a Generative RegIon-to-Text transformer, GRiT, for object understanding. The spirit of GRiT is to formulate object understanding as <region, text> pairs, where region locates objects and text describes objects. Specifically, GRiT consists of a visual encoder to extract image features, a foreground object extractor to localize objects, and a text decoder to generate natural language for objects. With the same model architecture, GRiT describes objects via not only simple nouns, but also rich descriptive sentences. We define GRiT as open-set object understanding, as it has no limit on object description output from the model architecture perspective. Experimentally, we apply GRiT to dense captioning and object detection tasks. GRiT achieves superior dense captioning performance (15.5 mAP on Visual Genome) and competitive detection accuracy (60.4 AP on COCO test-dev). Code is available at https://github.com/JialianW/GRiT.

Original languageEnglish
Title of host publicationComputer Vision – ECCV 2024 - 18th European Conference, Proceedings
EditorsAleš Leonardis, Elisa Ricci, Stefan Roth, Olga Russakovsky, Torsten Sattler, Gül Varol
PublisherSpringer Science and Business Media Deutschland GmbH
Pages207-224
Number of pages18
ISBN (Print)9783031729881
DOIs
StatePublished - 2025
Event18th European Conference on Computer Vision, ECCV 2024 - Milan, Italy
Duration: Sep 29 2024Oct 4 2024

Publication series

NameLecture Notes in Computer Science
Volume15138 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference18th European Conference on Computer Vision, ECCV 2024
Country/TerritoryItaly
CityMilan
Period09/29/2410/4/24

Fingerprint

Dive into the research topics of 'GRiT: A Generative Region-to-Text Transformer for Object Understanding'. Together they form a unique fingerprint.

Cite this