Machine Intelligence that Understands Visual and Linguistic Information and Interacts with Humans and Environments

Nguyen Van Quang · 2026

Read Machine Intelligence that Understands Visual and Linguistic Information and Interacts with Humans and Environments by Nguyen Van Quang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-24020.txt.

Introduction

Nguyen's dissertation advances vision-language intelligence across captions, visual dialog, and interactive instruction following. For captions, GRIT merges grid and region cues through a DETR-based detector, trains end to end, and beats prior methods on scores and speed. For dialog, LTMI, a light-weight transformer for many inputs, matches a standard transformer version with under one-tenth of its weights on VisDial. For instruction following on ALFRED, a two-stage setup first decodes orders without visual context, then fuses visual cues, reaching an unseen success rate of 8.37 percent.

Catalog record

CREATORNguyen Van Quang
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE200123
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.24020

Browse the full library · About the Tower of Babel