Fast Interactive Object Annotation With Curve-GCN

Ling, Huan; Gao, Jun; Kar, Amlan; Chen, Wenzheng; Fidler, Sanja

doi:10.1109/cvpr.2019.00540

Cited by 260 publications

(199 citation statements)

References 36 publications

(83 reference statements)

Supporting

Mentioning

198

Contrasting

Unclassified

Order By: Relevance

“…Note that these likely contain very different amounts of information that are relevant for recognition. For example, one may need to look at the complete and detailed object boundary to get a discriminative encoding of shape [2,33], while color and texture contain fairly low-level information. This may also provide an insight of why residual [19], skip [19,52] or even dense connections [21] lead to the most prominent performance gains.…”

Section: Introductionmentioning

confidence: 99%

Gated-SCNN: Gated Shape CNNs for Semantic Segmentation

Takikawa

Acuna

Jampani

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

Self Cite

628

391

View full text Add to dashboard Cite

Current state-of-the-art methods for image segmentation form a dense image representation where the color, shape and texture information are all processed together inside a deep CNN. This however may not be ideal as they contain very different type of information relevant for recognition. Here, we propose a new two-stream CNN architecture for semantic segmentation that explicitly wires shape information as a separate processing branch, i.e. shape stream, that processes information in parallel to the classical stream. Key to this architecture is a new type of gates that connect the intermediate layers of the two streams. Specifically, we use the higher-level activations in the classical stream to gate the lower-level activations in the shape stream, effectively removing noise and helping the shape stream to only focus on processing the relevant boundary-related information. This enables us to use a very shallow architecture for the shape stream that operates on the image-level resolution. Our experiments show that this leads to a highly effective architecture that produces sharper predictions around object boundaries and significantly boosts performance on thinner and smaller objects. Our method achieves state-ofthe-art performance on the Cityscapes benchmark, in terms of both mask (mIoU) and boundary (F-score) quality, improving by 2% and 4% over strong baselines.

show abstract

Section: Introductionmentioning

confidence: 99%

Gated-SCNN: Gated Shape CNNs for Semantic Segmentation

Takikawa

Acuna

Jampani

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

Self Cite

628

391

View full text Add to dashboard Cite

show abstract

“…For example, rather than representing a polygon as a discretization of some continuous function, [4,2] use a recurrent neural network (RNN) to directly predict the polygon vertices in a sequential manner. In [12], the authors predict the polygon or spline outlining the object using a Graph Convolutional Network.…”

Section: Related Workmentioning

confidence: 99%

DARNet: Deep Active Ray Network for Building Segmentation

Cheng

Liao

Fidler

et al. 2019

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

Self Cite

117

View full text Add to dashboard Cite

In this paper, we propose a Deep Active Ray Network (DARNet) for automatic building segmentation. Taking an image as input, it first exploits a deep convolutional neural network (CNN) as the backbone to predict energy maps, which are further utilized to construct an energy function. A polygon-based contour is then evolved via minimizing the energy function, of which the minimum defines the final segmentation. Instead of parameterizing the contour using Euclidean coordinates, we adopt polar coordinates, i.e., rays, which not only prevents self-intersection but also simplifies the design of the energy function. Moreover, we propose a loss function that directly encourages the contours to match building boundaries. Our DARNet is trained end-to-end by back-propagating through the energy minimization and the backbone CNN, which makes the CNN adapt to the dynamics of the contour evolution. Experiments on three building instance segmentation datasets demonstrate our DAR-Net achieves either state-of-the-art or comparable performances to other competitors.

show abstract

“…By uniformly sampling N points on (1), we then retrieve B, an equally-sized set of points describing the predicted contour. Inspired by the work of [20] on image annotation, we train our network with the loss…”

Section: Loss Functionmentioning

confidence: 99%

A Learning-Based Formulation of Parametric Curve Fitting for Bioimage Analysis

Mandal

Uhlmann

2020

Preprint

View full text Add to dashboard Cite

Parametric curve models are convenient to describe and quantitatively characterize the contour of objects in bioimages. Unfortunately, designing algorithms to fit smoothly such models onto image data classically requires significant domain expertise. Here, we propose a convolutional neural network-based approach to predict a continuous parametric representation of the outline of biological objects. We successfully apply our method on the Kaggle 2018 Data Science Bowl dataset composed of a varied collection of images of cell nuclei. This work is a first step towards user-friendly bioimage analysis tools that extract continuously-defined representations of objects.

show abstract

Fast Interactive Object Annotation With Curve-GCN

Cited by 260 publications

References 36 publications

Gated-SCNN: Gated Shape CNNs for Semantic Segmentation

Gated-SCNN: Gated Shape CNNs for Semantic Segmentation

DARNet: Deep Active Ray Network for Building Segmentation

A Learning-Based Formulation of Parametric Curve Fitting for Bioimage Analysis

Contact Info

Product

Resources

About