Interactive Image Segmentation with Latent Diversity

Li, Zhuwen; Chen, Qifeng; Koltun, Vladlen

doi:10.1109/cvpr.2018.00067

Cited by 197 publications

(163 citation statements)

References 28 publications

Supporting

Mentioning

163

Contrasting

Order By: Relevance

“…structure for both networks f and g, and modify the architecture to fit our purpose. We add a 1 × 1 convolutional layer at the beginning of each network to reduce the dimensionality of the input augmented with hypercolumn features [19]. To compute the optical flow, we use the state-of-theart method PWC-Net [28].…”

Section: Methodsmentioning

confidence: 99%

“…However, it is difficult to sample densely in the two-dimensional chrominance with hundreds of points. Thus we propose to use a perceptual loss with diversity [19] to avoid the discretization problem.…”

Section: Related Workmentioning

confidence: 99%

“…Generating a diverse set of solutions can be an effective way to tackle this multi-modality challenge. Inspired by the ranked diversity loss proposed by Li et al [19], we propose to generate multiple colorized videos to differentiate different solution modes. Besides, the diversity loss also contributes a lot to the temporal coherence because it reduces the ambiguity of colorization by generating several modes.…”

Section: Diverse Colorizationmentioning

confidence: 99%

“…By enforcing pairwise similarity between pixel pairs, we can preserve spatiotemporal color consistency in a video. Our model is also capable of generating multiple diverse colorized videos with a diversity loss [19]. We further suggest a simple strategy to select the most colorful video among all colorized videos.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Fully Automatic Video Colorization With Self-Regularization and Diversity

Lei¹,

Chen²

2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Self Cite

111

View full text Add to dashboard Cite

We present a fully automatic approach to video colorization with self-regularization and diversity. Our model contains a colorization network for video frame colorization and a refinement network for spatiotemporal color refinement. Without any labeled data, both networks can be trained with self-regularized losses defined in bilateral and temporal space. The bilateral loss enforces color consistency between neighboring pixels in a bilateral space and the temporal loss imposes constraints between corresponding pixels in two nearby frames. While video colorization is a multi-modal problem, our method uses a perceptual loss with diversity to differentiate various modes in the solution space. Perceptual experiments demonstrate that our approach outperforms state-of-the-art approaches on fully automatic video colorization.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Diverse Colorizationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Fully Automatic Video Colorization With Self-Regularization and Diversity

Lei¹,

Chen²

2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Self Cite

111

View full text Add to dashboard Cite

show abstract

“…Our work is novel in that we evaluate an interactive corrective annotation procedure in terms of annotation time to reach a certain accuracy on real-world plant image datasets. Synthetic data is often used to evaluate interactive segmentation methods (29)(30)(31). To provide more realistic measurements of annotation time we use real human annotators for our experiments.…”

Section: Introductionmentioning

confidence: 99%

RootPainter: Deep Learning Segmentation of Biological Images with Corrective Annotation

Smith

Han

Petersen

et al. 2020

Preprint

View full text Add to dashboard Cite

We present RootPainter, a GUI-based software tool for the rapid training of deep neural networks for use in biological image analysis. RootPainter facilitates both fully-automatic and semiautomatic image segmentation. We investigate the effectiveness of RootPainter using three plant image datasets, evaluating its potential for root length extraction from chicory roots in soil, biopore counting and root nodule counting from scanned roots. We also use RootPainter to compare dense annotations to corrective ones which are added during the training based on the weaknesses of the current model. Deep Learning | GUI | Segmentation | Phenotyping | Biopore | Rhizotron | Root nodule | Interactive segmentation Correspondence: ags@di.ku.dk Fig. 1. RootPainter corrective annotation concept. (a) Roots in soil. (b) AI root predictions. (c) Human corrections. (d) AI learns from corrections. Smith et al. | bioRχiv | April 16, 2020 | 1-16

show abstract

TriClick: Interactive Dermoscopic Image Segmentation with Triangle Map

Zhao,

Gu,

et al. 2024

IEEJ Transactions Elec Engng

View full text Add to dashboard Cite

Lesion segmentation is a fundamental task that has been widely studied in biomedicine. State‐of‐the‐art methods still struggle with the inherent challenges of dermoscopic images such as uncertain boundaries or severe occlusion. In this study, we argue that this task can be better resolved by introducing user interaction and proposing a deep interactive framework for lesion segmentation. Our method is designed to best reflect user expertise to iteratively refine the segmentation results. In each iteration, professional users are expected to click three points on the boundary of the lesion to encourage the network to update the previous round of the segmentation result. To best exploit the clues from the interaction, we explore a novel encoding strategy by producing a triangle map whose vertices are the three clicked points. Owing to the strong guidance within such triangle maps, our method can outperform state‐of‐the‐art methods with baseline segmentation backbones. Extensive experimental results and ablation studies on three commonly used lesion segmentation data sets demonstrated the effectiveness of our method in terms of segmentation accuracy and encoding effectiveness. © 2024 The Authors. IEEJ Transactions on Electrical and Electronic Engineering published by Institute of Electrical Engineer of Japan and Wiley Periodicals LLC.

show abstract

Interactive Image Segmentation with Latent Diversity

Cited by 197 publications

References 28 publications

Fully Automatic Video Colorization With Self-Regularization and Diversity

Fully Automatic Video Colorization With Self-Regularization and Diversity

RootPainter: Deep Learning Segmentation of Biological Images with Corrective Annotation

TriClick: Interactive Dermoscopic Image Segmentation with Triangle Map

Contact Info

Product

Resources

About