Repository logo
 

VideoNavQA: Bridging the Gap between Visual and Embodied Question Answering

Accepted version
Peer-reviewed

Type

Conference Object

Change log

Authors

Cangea, Cătălina 
Belilovsky, Eugene 
Liò, Pietro 
Courville, Aaron 

Abstract

Embodied Question Answering (EQA) is a recently proposed task, where an agent is placed in a rich 3D environment and must act based solely on its egocentric input to answer a given question. The desired outcome is that the agent learns to combine capabilities such as scene understanding, navigation and language understanding in order to perform complex reasoning in the visual world. However, initial advancements combining standard vision and language methods with imitation and reinforcement learning algorithms have shown EQA might be too complex and challenging for these techniques. In order to investigate the feasibility of EQA-type tasks, we build the VideoNavQA dataset that contains pairs of questions and videos generated in the House3D environment. The goal of this dataset is to assess question-answering performance from nearly-ideal navigation paths, while considering a much more complete variety of questions than current instantiations of the EQA task. We investigate several models, adapted from popular VQA methods, on this new benchmark. This establishes an initial understanding of how well VQA-style methods can perform within this novel EQA paradigm.

Description

Keywords

cs.CV, cs.CV, cs.AI, cs.CL, cs.LG

Journal Title

CoRR

Conference Name

The British Machine Vision Conference (BMVC) 2019

Journal ISSN

Volume Title

Publisher

Sponsorship
CC is funded by DREAM CDT and was supported by Mila during the time in Montréal. EB is funded by IVADO. We also thank the University of Cambridge Research Computing Services for providing HPC cluster resources.