# Learning on the Edge: Explicit Boundary Handling in CNNs

arXiv: https://arxiv.org/abs/1805.03106 · HTML: https://mlsanity.com/paper/1805.03106/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 8 May 2018.

## Built on by (most cited first, 6 of 6)

- [Padding Module: Learning the Padding in Deep Neural Networks](https://mlsanity.com/paper/2301.04608/) (2023)
- [Beyond Fine-tuning: Classifying High Resolution Mammograms using Function-Preserving Transformations](https://mlsanity.com/paper/2101.07945/) (2021)
- [Multifidelity data fusion in convolutional encoder/decoder networks](https://mlsanity.com/paper/2205.05187/) (2022)
- [Padding-free Convolution based on Preservation of Differential Characteristics of Kernels](https://mlsanity.com/paper/2309.06370/) (2023)
- [Improving Translation Invariance in Convolutional Neural Networks with Peripheral Prediction Padding](https://mlsanity.com/paper/2307.07725/) (2023)
- [Overcoming Labelled Data Scarcity for Defect Classification in Scanning Tunneling Microscopy](https://mlsanity.com/paper/2506.01678/) (2025)

## Abstract (arXiv)

Convolutional neural networks (CNNs) handle the case where filters extend beyond the image boundary using several heuristics, such as zero, repeat or mean padding. These schemes are applied in an ad-hoc fashion and, being weakly related to the image content and oblivious of the target task, result in low output quality at the boundary. In this paper, we propose a simple and effective improvement that learns the boundary handling itself. At training-time, the network is provided with a separate set of explicit boundary filters. At testing-time, we use these filters which have learned to extrapolate features at the boundary in an optimal way for the specific task. Our extensive evaluation, over a wide range of architectural changes (variations of layers, feature channels, or both), shows how the explicit filters result in improved boundary handling. Consequently, we demonstrate an improvement of 5% to 20% across the board of typical CNN applications (colorization, de-Bayering, optical flow, and disparity estimation).