# BoundaryNet: An Attentive Deep Network with Fast Marching Distance Maps for Semi-automatic Layout Annotation

arXiv: https://arxiv.org/abs/2108.09433 · HTML: https://mlsanity.com/paper/2108.09433/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 43 references that are in the map.
- First posted on arXiv 21 Aug 2021.

## Builds on (most cited first, 25 of 25)

- [Semi-Supervised Classification with Graph Convolutional Networks](https://mlsanity.com/paper/1609.02907/) (2016)
- [Mask R-CNN](https://mlsanity.com/paper/1703.06870/) (2017)
- [SGDR: Stochastic Gradient Descent with Warm Restarts](https://mlsanity.com/paper/1608.03983/) (2016)
- [Attention U-Net: Learning Where to Look for the Pancreas](https://mlsanity.com/paper/1804.03999/) (2018)
- [Variational Graph Auto-Encoders](https://mlsanity.com/paper/1611.07308/) (2016)
- [Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition](https://mlsanity.com/paper/1406.4729/) (2014)
- [DeepGCNs: Can GCNs Go as Deep as CNNs?](https://mlsanity.com/paper/1904.03751/) (2019)
- [PubLayNet: largest dataset ever for document layout analysis](https://mlsanity.com/paper/1908.07836/) (2019)
- [Evaluation of Deep Convolutional Nets for Document Image Classification and Retrieval](https://mlsanity.com/paper/1502.07058/) (2015)
- [On Regularized Losses for Weakly-supervised CNN Segmentation](https://mlsanity.com/paper/1803.09569/) (2018)
- [Efficient Interactive Annotation of Segmentation Datasets with Polygon-RNN++](https://mlsanity.com/paper/1803.09693/) (2018)
- [Box-driven Class-wise Region Masking and Filling Rate Guided Loss for Weakly Supervised Semantic Segmentation](https://mlsanity.com/paper/1904.11693/) (2019)
- [Fast Interactive Object Annotation with Curve-GCN](https://mlsanity.com/paper/1903.06874/) (2019)
- [Learning to Extract Semantic Structure from Documents Using Multimodal Fully Convolutional Neural Network](https://mlsanity.com/paper/1706.02337/) (2017)
- [Learning to Segment Medical Images with Scribble-Supervision Alone](https://mlsanity.com/paper/1807.04668/) (2018)
- [Learning deep structured active contours end-to-end](https://mlsanity.com/paper/1803.06329/) (2018)
- [Locating Objects Without Bounding Boxes](https://mlsanity.com/paper/1806.07564/) (2018)
- [Stacked Deconvolutional Network for Semantic Segmentation](https://mlsanity.com/paper/1708.04943/) (2017)
- [End to End Trainable Active Contours via Differentiable Rendering](https://mlsanity.com/paper/1912.00367/) (2019)
- [COCO_TS Dataset: Pixel-level Annotations Based on Weak Supervision for Scene Text Segmentation](https://mlsanity.com/paper/1904.00818/) (2019)
- [ICDAR 2019 Historical Document Reading Challenge on Large Structured Chinese Family Records](https://mlsanity.com/paper/1903.03341/) (2019)
- [Indiscapes: Instance Segmentation Networks for Layout Parsing of Historical Indic Manuscripts](https://mlsanity.com/paper/1912.07025/) (2019)
- [Learning Deep Representations for Word Spotting Under Weak Supervision](https://mlsanity.com/paper/1712.00250/) (2017)
- [Graph Node-Feature Convolution for Representation Learning](https://mlsanity.com/paper/1812.00086/) (2018)
- [Deep Learning Approaches for Image Retrieval and Pattern Spotting in Ancient Documents](https://mlsanity.com/paper/1907.09404/) (2019)

## Abstract (arXiv)

Precise boundary annotations of image regions can be crucial for downstream applications which rely on region-class semantics. Some document collections contain densely laid out, highly irregular and overlapping multi-class region instances with large range in aspect ratio. Fully automatic boundary estimation approaches tend to be data intensive, cannot handle variable-sized images and produce sub-optimal results for aforementioned images. To address these issues, we propose BoundaryNet, a novel resizing-free approach for high-precision semi-automatic layout annotation. The variable-sized user selected region of interest is first processed by an attention-guided skip network. The network optimization is guided via Fast Marching distance maps to obtain a good quality initial boundary estimate and an associated feature representation. These outputs are processed by a Residual Graph Convolution Network optimized using Hausdorff loss to obtain the final region boundary. Results on a challenging image manuscript dataset demonstrate that BoundaryNet outperforms strong baselines and produces high-quality semantic region boundaries. Qualitatively, our approach generalizes across multiple document image datasets containing different script systems and layouts, all without additional fine-tuning. We integrate BoundaryNet into a document annotation system and show that it provides high annotation throughput compared to manual and fully automatic alternatives.