# Domain-Size Pooling in Local Descriptors: DSP-SIFT

arXiv: https://arxiv.org/abs/1412.8556 · HTML: https://mlsanity.com/paper/1412.8556/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 30 Dec 2014.

## Built on by (most cited first, 14 of 14)

- [Parallel Structure from Motion for UAV Images via Weighted Connected Dominating Set](https://mlsanity.com/paper/2206.11499/) (2022)
- [SIFT Matching by Context Exposed](https://mlsanity.com/paper/2106.09584/) (2021)
- [Sub-token ViT Embedding via Stochastic Resonance Transformers](https://mlsanity.com/paper/2310.03967/) (2023)
- [Efficient Pairwise Neuroimage Analysis using the Soft Jaccard Index and 3D Keypoint Sets](https://mlsanity.com/paper/2103.06966/) (2021)
- [Feature matching for multi-epoch historical aerial images](https://mlsanity.com/paper/2112.04255/) (2021)
- [Semihierarchical Reconstruction and Weak-area Revisiting for Robotic Visual Seafloor Mapping](https://mlsanity.com/paper/2308.06147/) (2023)
- [VIO-Aided Structure from Motion Under Challenging Environments](https://mlsanity.com/paper/2101.09657/) (2021)
- [GPU optimization of the 3D Scale-invariant Feature Transform Algorithm and a Novel BRIEF-inspired 3D Fast Descriptor](https://mlsanity.com/paper/2112.10258/) (2021)
- [Descriptor Distillation: a Teacher-Student-Regularized Framework for Learning Local Descriptors](https://mlsanity.com/paper/2209.11795/) (2022)
- [Region Feature Descriptor Adapted to High Affine Transformations](https://mlsanity.com/paper/2402.09724/) (2024)
- [Cross Spatial Temporal Fusion Attention for Remote Sensing Object Detection via Image Feature Matching](https://mlsanity.com/paper/2507.19118/) (2025)
- [Exploring Image Representation with Decoupled Classical Visual Descriptors](https://mlsanity.com/paper/2510.14536/) (2025)
- [PySIFT: GPU-Resident Deterministic SIFT for Deep Learning Vision Pipelines](https://mlsanity.com/paper/2605.17869/) (2026)
- [Evaluating Mesh Reconstruction Methods for Crop Phenotyping](https://mlsanity.com/paper/2609.16926/) (2026)

## Abstract (arXiv)

We introduce a simple modification of local image descriptors, such as SIFT, based on pooling gradient orientations across different domain sizes, in addition to spatial locations. The resulting descriptor, which we call DSP-SIFT, outperforms other methods in wide-baseline matching benchmarks, including those based on convolutional neural networks, despite having the same dimension of SIFT and requiring no training.