# An Efficient Approximate kNN Graph Method for Diffusion on Image Retrieval

## TL;DR

This paper introduces a novel LSH-based approximate kNN graph method that significantly accelerates diffusion processes in image retrieval, maintaining performance while reducing computation time by approximately 18 times.

## Contribution

The paper presents an efficient approximate kNN graph construction technique using LSH projections that matches exact diffusion performance with much faster computation.

## Key findings

- Achieves similar diffusion performance as exact kNN graphs.
- Approximately 18 times faster on large image datasets.
- Validated on multiple public image datasets.

## Abstract

The application of the diffusion in many computer vision and artificial intelligence projects has been shown to give excellent improvements in performance. One of the main bottlenecks of this technique is the quadratic growth of the kNN graph size due to the high-quantity of new connections between nodes in the graph, resulting in long computation times. Several strategies have been proposed to address this, but none are effective and efficient. Our novel technique, based on LSH projections, obtains the same performance as the exact kNN graph after diffusion, but in less time (approximately 18 times faster on a dataset of a hundred thousand images). The proposed method was validated and compared with other state-of-the-art on several public image datasets, including Oxford5k, Paris6k, and Oxford105k.