Is Two-shot All You Need? A Label-efficient Approach for Video Segmentation in Breast Ultrasound

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Zeng, Jiajun, Ni, Dong, Huang, Ruobing
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866916145619009536
author Zeng, Jiajun
Ni, Dong
Huang, Ruobing
author_facet Zeng, Jiajun
Ni, Dong
Huang, Ruobing
contents Breast lesion segmentation from breast ultrasound (BUS) videos could assist in early diagnosis and treatment. Existing video object segmentation (VOS) methods usually require dense annotation, which is often inaccessible for medical datasets. Furthermore, they suffer from accumulative errors and a lack of explicit space-time awareness. In this work, we propose a novel two-shot training paradigm for BUS video segmentation. It not only is able to capture free-range space-time consistency but also utilizes a source-dependent augmentation scheme. This label-efficient learning framework is validated on a challenging in-house BUS video dataset. Results showed that it gained comparable performance to the fully annotated ones given only 1.9% training labels.
format Preprint
id arxiv_https___arxiv_org_abs_2402_04921
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Is Two-shot All You Need? A Label-efficient Approach for Video Segmentation in Breast Ultrasound
Zeng, Jiajun
Ni, Dong
Huang, Ruobing
Image and Video Processing
Computer Vision and Pattern Recognition
I.4.6
Breast lesion segmentation from breast ultrasound (BUS) videos could assist in early diagnosis and treatment. Existing video object segmentation (VOS) methods usually require dense annotation, which is often inaccessible for medical datasets. Furthermore, they suffer from accumulative errors and a lack of explicit space-time awareness. In this work, we propose a novel two-shot training paradigm for BUS video segmentation. It not only is able to capture free-range space-time consistency but also utilizes a source-dependent augmentation scheme. This label-efficient learning framework is validated on a challenging in-house BUS video dataset. Results showed that it gained comparable performance to the fully annotated ones given only 1.9% training labels.
title Is Two-shot All You Need? A Label-efficient Approach for Video Segmentation in Breast Ultrasound
topic Image and Video Processing
Computer Vision and Pattern Recognition
I.4.6
url https://arxiv.org/abs/2402.04921