HCT-net: hybrid CNN-transformer model based on a neural architecture search network for medical image segmentation

Yu, Zhihong; Lee, Feifei; Chen, Qiu

doi:10.1007/s10489-023-04570-z

Cited by 6 publications

(1 citation statement)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These categories include online fine-tuning-based methods, motion-based methods and matching-based methods. Online fine-tuning-based methods, exemplified by works such as OSVOS [29], OSVOS-S [30] and OnAVOS [38], based on the idea of image segmentation [49][50][51], involve the fine-tuning of general segmentation networks on the initial annotated frame to target specific objects. In order to address the evolving appearance of objects over time, these methods frequently conduct online updates of their segmentation models.…”

Section: Related Workmentioning

confidence: 99%

Modulated Memory Network for Video Object Segmentation

Lu,

Guo,

Zuo

2024

Mathematics

View full text Add to dashboard Cite

Existing video object segmentation (VOS) methods based on matching techniques commonly employ a reference set comprising historical segmented frames, referred to as ‘memory frames’, to facilitate the segmentation process. However, these methods suffer from the following limitations: (i) Inherent segmentation errors in memory frames can propagate and accumulate errors when utilized as templates for subsequent segmentation. (ii) The non-local matching technique employed in top-leading solutions often fails to incorporate positional information, potentially leading to incorrect matching. In this paper, we introduce the Modulated Memory Network (MMN) for VOS. Our MMN enhances matching-based VOS methods in the following ways: (i) Introducing an Importance Modulator, which adjusts memory frames using adaptive weight maps generated based on the segmentation confidence associated with each frame. (ii) Incorporating a Position Modulator that encodes spatial and temporal positional information for both memory frames and the current frame. The proposed modulator improves matching accuracy by embedding positional information. Meanwhile, the Importance Modulator mitigates error propagation and accumulation by incorporating confidence-based modulation. Through extensive experimentation, we demonstrate the effectiveness of our proposed MMN, which also achieves promising performance on VOS benchmarks.

show abstract