# Improving the HardNet Descriptor

arXiv: https://arxiv.org/abs/2007.09699 · HTML: https://mlsanity.com/paper/2007.09699/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 19 Jul 2020.

## Built on by (most cited first, 4 of 4)

- [COTR: Correspondence Transformer for Matching Across Images](https://mlsanity.com/paper/2103.14167/) (2021)
- [HarrisZ^+: Harris Corner Selection for Next-Gen Image Matching Pipelines](https://mlsanity.com/paper/2109.12925/) (2021)
- [A2B: Anchor to Barycentric Coordinate for Robust Correspondence](https://mlsanity.com/paper/2306.02760/) (2023)
- [Tiny CNN for feature point description for document analysis: approach and dataset](https://mlsanity.com/paper/2109.04134/) (2021)

## Abstract (arXiv)

In the thesis we consider the problem of local feature descriptor learning for wide baseline stereo focusing on the HardNet descriptor, which is close to state-of-the-art. AMOS Patches dataset is introduced, which improves robustness to illumination and appearance changes. It is based on registered images from selected cameras from the AMOS dataset. We provide recommendations on the patch dataset creation process and evaluate HardNet trained on data of different modalities. We also introduce a dataset combination and reduction methods, that allow comparable performance on a significantly smaller dataset. HardNet8, consistently outperforming the original HardNet, benefits from the architectural choices made: connectivity pattern, final pooling, receptive field, CNN building blocks found by manual or automatic search algorithms -- DARTS. We show impact of overlooked hyperparameters such as batch size and length of training on the descriptor quality. PCA dimensionality reduction further boosts performance and also reduces memory footprint. Finally, the insights gained lead to two HardNet8 descriptors: one performing well on a variety of benchmarks -- HPatches, AMOS Patches and IMW Phototourism, the other is optimized for IMW Phototourism.