Multi-level U-net network for image super-resolution reconstruction

Han, Ning; Zhou, Li; Xie, Zhengmao; Zheng, Jingli; Zhang, Liuxin

doi:10.1016/j.displa.2022.102192

Cited by 28 publications

(4 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

Section: Methodsmentioning

confidence: 99%

“…This design facilitates the incorporation of both global and local information, making it particularly effective for tasks where accurate delineation of structures is crucial, such as in identifying organs [40,41], tumors [42], and anatomical features [43,39,44]. Beyond image segmentation, the U-Net's versatility has led to its adoption in various medical imaging applications, including image denoising [45,46], registration [47,48], and super-resolution [49,50,51], showcasing its adaptability and robust performance across different scenarios. Multi-resolution U-Nets: To overcome the limitations related to the size of the data set and sparse annotations described in the introduction, we propose a cascaded resolution approach, inspired by previous works [29,52], in combination with semi-supervised learning, which takes in volumetric inputs downsampled at different resolutions, while ensuring that all U-Net components receive inputs of the same size.…”

Section: Semi-supervised Segmentation Modelmentioning

confidence: 99%

See 1 more Smart Citation

Segmentation of supragranular and infragranular layers in ultra-high resolution 7Tex vivoMRI of the human cerebral cortex

Zeng,

Puonti,

Sayeed

et al. 2023

Preprint

View full text Add to dashboard Cite

Accurate labeling of specific layers in the human cerebral cortex is crucial for advancing our understanding of neurodevelopmental and neurodegenerative disorders. Lever-aging recent advancements in ultra-high resolutionex vivoMRI, we present a novel semi-supervised segmentation model capable of identifying supragranular and infragranular layers inex vivoMRI with unprecedented precision. On a dataset consisting of 17 whole-hemisphereex vivoscans at 120µm, we propose a multi-resolution U-Nets framework (MUS) that integrates global and local structural information, achieving reliable segmentation maps of the entire hemisphere, with Dice scores over 0.8 for supra- and infragranular layers. This enables surface modeling, atlas construction, anomaly detection in disease states, and cross-modality validation, while also paving the way for finer layer segmentation. Our approach offers a powerful tool for comprehensive neuroanatomical investigations and holds promise for advancing our mechanistic understanding of progression of neurodegenerative diseases.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Semi-supervised Segmentation Modelmentioning

confidence: 99%

Segmentation of supragranular and infragranular layers in ultra-high resolution 7Tex vivoMRI of the human cerebral cortex

Zeng,

Puonti,

Sayeed

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Although the feature extraction module in STTR can extract multiscale features of images, the information in these feature maps is not fully utilized, resulting in a serious mismatching phenomenon. DispNet [16] adopts a network structure similar to U-net [17] for feature extraction, which effectively fuses local and global information. However, the relationship between feature maps of the same scale in DispNet is not close, and the information among the feature maps is underutilized, thereby affecting the matching performance.…”

Section: Feature Extraction Layermentioning

confidence: 99%

“…DispNet [16] is the first end-to-end stereo matching method using deep learning. DispNet [16] adopts a network structure similar to Unet [17] for feature extraction, leveraging CNNs to extract multiscale feature maps and fuse local and global information. However, the weak relationship between feature maps in DispNet hinders effective information fusion, adversely affecting the matching process.…”

Section: Introductionmentioning

confidence: 99%

End-to-end information fusion method for transformer-based stereo matching

Xu,

Wang,

Guo

2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

In stereo matching, the application of transformers can overcome the limitations of disparity range and capture long-range matching information. However, the lack of cross-epipolar context information often leads to numerous mismatches, especially in low-texture regions. An end-to-end information fusion stereo matching method is proposed to address this issue. In the proposed method, a feature extraction method that combines dense connections and a residual block is proposed. Global and local semantic information can be effectively fused by incorporating dense connections among multiscale feature maps. Additionally, the inclusion of a residual block helps extract more representative feature maps. The idea of criss-cross attention is introduced in the transformer implicit matching process. Criss-cross attention enables the capture of cross-epipolar context information by combining horizontal and vertical attention mechanisms. This method improves the matching accuracy from the perspective of multi-path information fusion. According to the matching resluts, the disparity regression layer and the context adjustment layer are used to generate the initial and final disparity maps, respectively. The proposed method is evaluated on the Scene Flow, KITTI 2012, and Middlebury 2014 datasets. Experimental results indicate that the proposed method effectively enhances matching accuracy. Moreover, the proposed method exhibits strong generalization ability, allowing for direct application to synthetic, real outdoor, and real indoor scene images.

show abstract