Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Goletto, Gabriele, Nagarajan, Tushar, Averta, Giuseppe, Damen, Dima
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2409.10917
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866910607155920896
author Goletto, Gabriele
Nagarajan, Tushar
Averta, Giuseppe
Damen, Dima
author_facet Goletto, Gabriele
Nagarajan, Tushar
Averta, Giuseppe
Damen, Dima
contents Egocentric videos provide a unique perspective into individuals' daily experiences, yet their unstructured nature presents challenges for perception. In this paper, we introduce AMEGO, a novel approach aimed at enhancing the comprehension of very-long egocentric videos. Inspired by the human's ability to maintain information from a single watching, AMEGO focuses on constructing a self-contained representations from one egocentric video, capturing key locations and object interactions. This representation is semantic-free and facilitates multiple queries without the need to reprocess the entire visual content. Additionally, to evaluate our understanding of very-long egocentric videos, we introduce the new Active Memories Benchmark (AMB), composed of more than 20K of highly challenging visual queries from EPIC-KITCHENS. These queries cover different levels of video reasoning (sequencing, concurrency and temporal grounding) to assess detailed video understanding capabilities. We showcase improved performance of AMEGO on AMB, surpassing other video QA baselines by a substantial margin.
format Preprint
id arxiv_https___arxiv_org_abs_2409_10917
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle AMEGO: Active Memory from long EGOcentric videos
Goletto, Gabriele
Nagarajan, Tushar
Averta, Giuseppe
Damen, Dima
Computer Vision and Pattern Recognition
Egocentric videos provide a unique perspective into individuals' daily experiences, yet their unstructured nature presents challenges for perception. In this paper, we introduce AMEGO, a novel approach aimed at enhancing the comprehension of very-long egocentric videos. Inspired by the human's ability to maintain information from a single watching, AMEGO focuses on constructing a self-contained representations from one egocentric video, capturing key locations and object interactions. This representation is semantic-free and facilitates multiple queries without the need to reprocess the entire visual content. Additionally, to evaluate our understanding of very-long egocentric videos, we introduce the new Active Memories Benchmark (AMB), composed of more than 20K of highly challenging visual queries from EPIC-KITCHENS. These queries cover different levels of video reasoning (sequencing, concurrency and temporal grounding) to assess detailed video understanding capabilities. We showcase improved performance of AMEGO on AMB, surpassing other video QA baselines by a substantial margin.
title AMEGO: Active Memory from long EGOcentric videos
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2409.10917