Superpixel Convolution for Segmentation

Suzuki, Teppei; Akizuki, Shuichi; Kato, Naoki; Aoki, Yoshimitsu

doi:10.1109/icip.2018.8451721

Cited by 7 publications

(8 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a detail-preserving complexity reduction for image data, superpixel segmentation is classically used, which groups pixels similar in color and other low-level properties. Superpixels can preserve object boundaries and semantics, and some existing methods [10,20,21,41,46,49] combine superpixels with deep neural networks. Many of them explicitly use superpixels as a downsampling operation and need special operations such as graph convolution to process the downsampled image because of the irregularity of superpixels or require an additional model to compute superpixels before downsampling.…”

Section: Related Workmentioning

confidence: 99%

“…Classically, superpixels [38] are utilized as an efficient image representation, which can reduce the complexity of image data and preserve the local structure. Some existing methods [21,41,46] combine superpixels with deep neural networks to reduce computational costs through the use of superpixel-based downsampling. Due to the irregularity of superpixels, they require graph convolution [20,30,41] to process the downsampled features.…”

Section: Introductionmentioning

confidence: 99%

“…Some existing methods [21,41,46] combine superpixels with deep neural networks to reduce computational costs through the use of superpixel-based downsampling. Due to the irregularity of superpixels, they require graph convolution [20,30,41] to process the downsampled features. Therefore, one needs to modify the feed-forward path of existing architectures and/or require an extra model to compute superpixels before downsampling.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Implicit Integration of Superpixel Segmentation into Fully Convolutional Networks

Suzuki¹

2021

Preprint

Self Cite

View full text Add to dashboard Cite

Superpixels are a useful representation to reduce the complexity of image data. However, to combine superpixels with convolutional neural networks (CNNs) in an end-toend fashion, one requires extra models to generate superpixels and special operations such as graph convolution. In this paper, we propose a way to implicitly integrate a superpixel scheme into CNNs, which makes it easy to use superpixels with CNNs in an end-to-end fashion. Our proposed method hierarchically groups pixels at downsampling layers and generates superpixels. Our method can be plugged into many existing architectures without a change in their feed-forward path because our method does not use superpixels in the feed-forward path but use them to recover the lost resolution instead of bilinear upsampling. As a result, our method preserves detailed information such as object boundaries in the form of superpixels even when the model contains downsampling layers. We evaluate our method on several tasks such as semantic segmentation, superpixel segmentation, and monocular depth estimation, and confirm that it speeds up modern architectures and/or improves their prediction accuracy in these tasks.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Implicit Integration of Superpixel Segmentation into Fully Convolutional Networks

Suzuki¹

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…One possible approach is to redesign a network architecture to allow the computation of irregular superpixel grids, e.g. [8,32,42]. This is still challenging, particularly, if one wishes to integrate subsequent tasks within a single learning process.…”

Section: Introductionmentioning

confidence: 99%

HERS Superpixels: Deep Affinity Learning for Hierarchical Entropy Rate Segmentation

Peng¹,

Avilés-Rivero²,

Schoenlieb

2022

2022 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)

View full text Add to dashboard Cite

Superpixels serve as a powerful preprocessing tool in many computer vision tasks. By using superpixel representation, the number of image primitives can be largely reduced by orders of magnitudes. The majority of superpixel methods use handcrafted features, which usually do not translate well into strong adherence to object boundaries. A few recent superpixel methods have introduced deep learning into the superpixel segmentation process. However, none of these methods is able to produce superpixels in near real-time, which is crucial to the applicability of a superpixel method in practice. In this work, we propose a two-stage graph-based framework for superpixel segmentation. In the first stage, we introduce an efficient Deep Affinity Learning (DAL) network that learns pairwise pixel affinities by aggregating multi-scale information. In the second stage, we propose a highly efficient superpixel method called Hierarchical Entropy Rate Segmentation (HERS). Using the learned affinities from the first stage, HERS builds a hierarchical tree structure that can produce any number of highly adaptive superpixels instantaneously. We demonstrate, through visual and numerical experiments, the effectiveness and efficiency of our method compared to various state-of-the-art superpixel methods.

show abstract

“…Comparing to pixels, superpixel provides a more effective representation for image data. With such a compact representation, the computational efficiency of vision algorithms could be improved [18,11,36]. Consequently, superpixel could benefit many vision tasks like…”

Section: Introductionmentioning

confidence: 99%

AINet: Association Implantation for Superpixel Segmentation

Wang

Wei²,

Qian

et al. 2021

Preprint

View full text Add to dashboard Cite

Recently, some approaches are proposed to harness deep convolutional networks to facilitate superpixel segmentation. The common practice is to first evenly divide the image into a pre-defined number of grids and then learn to associate each pixel with its surrounding grids. However, simply applying a series of convolution operations with limited receptive fields can only implicitly perceive the relations between the pixel and its surrounding grids. Consequently, existing methods often fail to provide an effective context when inferring the association map. To remedy this issue, we propose a novel Association Implantation (AI) module to enable the network to explicitly capture the relations between the pixel and its surrounding grids. The proposed AI module directly implants the features of grid cells to the surrounding of its corresponding central pixel, and conducts convolution on the padded window to adaptively transfer knowledge between them. With such an implantation operation, the network could explicitly harvest the pixel-grid level context, which is more in line with the target of superpixel segmentation comparing to the pixel-wise relation. Furthermore, to pursue better boundary precision, we design a boundary-perceiving loss to help the network discriminate the pixels around boundaries in hidden feature level, which could benefit the subsequent inferring modules to accurately identify more boundary pixels. Extensive experiments on BSDS500 and NYUv2 datasets show that our method could not only achieve state-of-the-art performance but maintain satisfactory inference efficiency.

show abstract

Superpixel Convolution for Segmentation

Cited by 7 publications

References 7 publications

Implicit Integration of Superpixel Segmentation into Fully Convolutional Networks

Implicit Integration of Superpixel Segmentation into Fully Convolutional Networks

HERS Superpixels: Deep Affinity Learning for Hierarchical Entropy Rate Segmentation

AINet: Association Implantation for Superpixel Segmentation

Contact Info

Product

Resources

About