Summary of Efficient Approximation Of Earth Mover’s Distance Based on Nearest Neighbor Search, by Guangyu Meng et al.
Efficient approximation of Earth Mover’s Distance Based on Nearest Neighbor Search
by Guangyu Meng, Ruyu Zhou, Liu Liu, Peixian Liang, Fang Liu, Danny Chen, Michael Niemier, X.Sharon Hu
First submitted to arxiv on: 14 Jan 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper presents a novel approach called NNS-EMD that efficiently approximates Earth Mover’s Distance (EMD), a crucial similarity measure in computer vision and other domains. The traditional EMD calculation is computationally and memory-intensive, hindering its scalability for large-scale problems. Existing approximate EMD algorithms trade accuracy for speed or require manual parameter tuning. NNS-EMD leverages Nearest Neighbor Search to reduce computational costs while maintaining high accuracy. This approach also enables parallel processing and vectorization on GPU, further accelerating the calculation. The authors compare NNS-EMD with exact EMD and state-of-the-art approximate methods on image classification and retrieval tasks, demonstrating superior accuracy, speedup, and memory efficiency. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper solves a problem in computer vision where a tool called Earth Mover’s Distance (EMD) is used to compare images. The traditional way of calculating EMD is slow and uses too much memory, making it hard to use for large datasets. The authors create a new method called NNS-EMD that quickly calculates EMD while keeping the accuracy high. This approach also lets computers work together in parallel, which helps even more. The paper compares this new method with others on image classification and retrieval tasks, showing that it’s faster and more efficient. |
Keywords
» Artificial intelligence » Image classification » Nearest neighbor » Vectorization
