# Stride Independent Patching for Deep Learning

arXiv: https://arxiv.org/abs/2610.12216 · HTML: https://mlsanity.com/paper/2610.12216/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 21 of its 37 references that are in the map.
- First posted on arXiv 8 Oct 2026.

## Builds on (most cited first, 21 of 21)

- [Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation](https://mlsanity.com/paper/1802.02611/) (2018)
- [Scaling Vision Transformers to Gigapixel Images via Hierarchical Self-Supervised Learning](https://mlsanity.com/paper/2206.02647/) (2022)
- [High-Resolution Image Inpainting using Multi-Scale Neural Patch Synthesis](https://mlsanity.com/paper/1611.09969/) (2016)
- [HookNet: multi-resolution convolutional neural networks for semantic segmentation in histopathology whole-slide images](https://mlsanity.com/paper/2006.12230/) (2020)
- [Streaming convolutional neural networks for end-to-end learning with multi-megapixel images](https://mlsanity.com/paper/1911.04432/) (2019)
- [FastSurferVINN: Building Resolution-Independence into Deep Learning Segmentation Methods -- A Solution for HighRes Brain MRI](https://mlsanity.com/paper/2112.09654/) (2021)
- [Introducing Hann windows for reducing edge-effects in patch-based image segmentation](https://mlsanity.com/paper/1910.07831/) (2019)
- [Advanced Steel Microstructural Classification by Deep Learning Methods](https://mlsanity.com/paper/1706.06480/) (2017)
- [Iterative Patch Selection for High-Resolution Image Recognition](https://mlsanity.com/paper/2210.13007/) (2022)
- [Evaluation of Rounding Functions in Nearest-Neighbor Interpolation](https://mlsanity.com/paper/2003.06885/) (2020)
- [Effects of Image Size on Deep Learning](https://mlsanity.com/paper/2101.11508/) (2021)
- [Adaptive Patching for High-resolution Image Segmentation with Transformers](https://mlsanity.com/paper/2404.09707/) (2024)
- [Software Implementation of Optimized Bicubic Interpolated Scan Conversion in Echocardiography](https://mlsanity.com/paper/2005.11269/) (2020)
- [Normalized Weighting Schemes for Image Interpolation Algorithms](https://mlsanity.com/paper/2011.08559/) (2020)
- [Effect of the regularization hyperparameter on deep learning-based segmentation in LGE-MRI](https://mlsanity.com/paper/2012.05661/) (2020)
- [Evaluation of Extra Pixel Interpolation with Mask Processing for Medical Image Segmentation with Deep Learning](https://mlsanity.com/paper/2302.11522/) (2023)
- [Convolutional Neural Networks for Automatic Detection of Intact Adenovirus from TEM Imaging with Debris, Broken and Artefacts Particles](https://mlsanity.com/paper/2310.19630/) (2023)
- [Advances on image interpolation based on ant colony algorithm](https://mlsanity.com/paper/2104.12863/) (2021)
- [Patch Gradient Descent: Training Neural Networks on Very Large Images](https://mlsanity.com/paper/2301.13817/) (2023)
- [Beyond Nearest Neighbor Interpolation in Data Augmentation](https://mlsanity.com/paper/2504.01527/) (2025)
- [Toward Optimal Adenovirus Detection Using YOLO26](https://mlsanity.com/paper/2607.17799/) (2026)

## Abstract (arXiv)

This paper presents semi-automatic stride-independent patching (SSP) as an alternative to automatic stride-dependent patching techniques. SSP uses user or expert input to position predefined patches over one or more objects of interest. To evaluate its effectiveness, three patch-based datasets were created using SSP, overlapping patching (Overlap), and non-overlapping patching (Noverlap). DeepLabV3+ models with ResNet50, ResNet18, and MobileNetV2 backbones were trained sepa-rately on each dataset. Quantitative evaluations were performed on the respective test splits and a common external test set. SSP generally achieved higher segmentation scores on the test splits and required the shortest model training time across all three backbones. On the external test set, SSP achieved the highest average precision and F1-score across backbones, whereas Noverlap achieved the highest average recall. These preliminary results demonstrate that the potentially greater spatial coverage of Noverlap and Overlap does not generally translate into better segmentation perfor-mance and that SSP offers a favorable balance between segmentation performance and model training time.