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

arXiv: https://arxiv.org/abs/2402.04921 · HTML: https://mlsanity.com/paper/2402.04921/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 8 of its 21 references that are in the map.
- First posted on arXiv 7 Feb 2024.

## Builds on (most cited first, 8 of 8)

- [XMem: Long-Term Video Object Segmentation with an Atkinson-Shiffrin Memory Model](https://mlsanity.com/paper/2207.07115/) (2022)
- [Video Object Segmentation using Space-Time Memory Networks](https://mlsanity.com/paper/1904.00607/) (2019)
- [ST++: Make Self-training Work Better for Semi-supervised Semantic Segmentation](https://mlsanity.com/paper/2106.05095/) (2021)
- [A Survey on Deep Learning Technique for Video Segmentation](https://mlsanity.com/paper/2107.01153/) (2021)
- [AAU-net: An Adaptive Attention U-net for Breast Lesions Segmentation in Ultrasound Images](https://mlsanity.com/paper/2204.12077/) (2022)
- [Boosting Video Object Segmentation via Space-time Correspondence Learning](https://mlsanity.com/paper/2304.06211/) (2023)
- [Shifting More Attention to Breast Lesion Segmentation in Ultrasound Videos](https://mlsanity.com/paper/2310.01861/) (2023)
- [Two-shot Video Object Segmentation](https://mlsanity.com/paper/2303.12078/) (2023)

## Abstract (arXiv)

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.