Improving Unsupervised Video Object Segmentation via Fake Flow Generation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Cho, Suhwan, Lee, Minhyeok, Lee, Jungho, Kim, Donghyeong, Lee, Seunghoon, Woo, Sungmin, Lee, Sangyoun
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909257545285632
author Cho, Suhwan
Lee, Minhyeok
Lee, Jungho
Kim, Donghyeong
Lee, Seunghoon
Woo, Sungmin
Lee, Sangyoun
author_facet Cho, Suhwan
Lee, Minhyeok
Lee, Jungho
Kim, Donghyeong
Lee, Seunghoon
Woo, Sungmin
Lee, Sangyoun
contents Unsupervised video object segmentation (VOS), also known as video salient object detection, aims to detect the most prominent object in a video at the pixel level. Recently, two-stream approaches that leverage both RGB images and optical flow maps have gained significant attention. However, the limited amount of training data remains a substantial challenge. In this study, we propose a novel data generation method that simulates fake optical flows from single images, thereby creating large-scale training data for stable network learning. Inspired by the observation that optical flow maps are highly dependent on depth maps, we generate fake optical flows by refining and augmenting the estimated depth maps of each image. By incorporating our simulated image-flow pairs, we achieve new state-of-the-art performance on all public benchmark datasets without relying on complex modules. We believe that our data generation method represents a potential breakthrough for future VOS research.
format Preprint
id arxiv_https___arxiv_org_abs_2407_11714
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Improving Unsupervised Video Object Segmentation via Fake Flow Generation
Cho, Suhwan
Lee, Minhyeok
Lee, Jungho
Kim, Donghyeong
Lee, Seunghoon
Woo, Sungmin
Lee, Sangyoun
Computer Vision and Pattern Recognition
Unsupervised video object segmentation (VOS), also known as video salient object detection, aims to detect the most prominent object in a video at the pixel level. Recently, two-stream approaches that leverage both RGB images and optical flow maps have gained significant attention. However, the limited amount of training data remains a substantial challenge. In this study, we propose a novel data generation method that simulates fake optical flows from single images, thereby creating large-scale training data for stable network learning. Inspired by the observation that optical flow maps are highly dependent on depth maps, we generate fake optical flows by refining and augmenting the estimated depth maps of each image. By incorporating our simulated image-flow pairs, we achieve new state-of-the-art performance on all public benchmark datasets without relying on complex modules. We believe that our data generation method represents a potential breakthrough for future VOS research.
title Improving Unsupervised Video Object Segmentation via Fake Flow Generation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2407.11714