Ultrafast Euclidean Shortest Path Computation Using Hub Labeling

Authors: Jinchun Du, Bojie Shen, Muhammad Aamir Cheema

AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our experimental study conducted on the widely used benchmark maps shows that our approach is typically 1-2 orders of magnitude faster than two state-of-the-art algorithms.
Researcher Affiliation Academia Faculty of Information Technology, Monash University, Melbourne, Australia {jinchun.du, bojie.shen, aamir.cheema}@monash.edu
Pseudocode Yes Algorithm 1 shows the details of how to compute vdistmin(p, hi). Algorithm 2 shows the details of our query processing algorithm.
Open Source Code Yes For reproducibility, implementation of EHL is available online4. 4https://github.com/goldi1027/EHL
Open Datasets Yes Similar to the existing studies, we conduct experiments on the widely used game map benchmarks2, described in (Sturtevant 2012a), on a total of 373 game maps (see Table 3). 2https://github.com/nathansttt/hog2
Dataset Splits No The paper does not explicitly provide training/test/validation dataset splits with specific percentages, sample counts, or references to predefined splits.
Hardware Specification Yes We run experiments on a 3.2 GHz Intel Core i7 machine with 32 GB of RAM.
Software Dependencies No All the algorithms are implemented in C++ and complied with -O3 flag. This only mentions the programming language and a compiler flag, not specific software components with version numbers.
Experiment Setup Yes We evaluate the effect of size of the uniform grid used in our Euclidean Hub Labeling (EHL) approach by using grids where each cell is N times bigger than the base cell in each dimension. We vary N from 1 to 64 (e.g., cell sizes vary from 1x to 64x of the base cell size in each dimension)... Ablation Study : In Table 5 and Table 6, we show the effectiveness of the four pruning rules and the optimisation (e.g., by removing these one at a time) on build time, memory and query performance.