Hashing for Distributed Data

Authors: Cong Leng, Jiaxiang Wu, Jian Cheng, Xi Zhang, Hanqing Lu

ICML 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on several large scale datasets containing up to 100 million samples demonstrate the efficacy of our method. and In order to evaluate the performance of the proposed distributed hashing (Dis H), we conduct a series of experiments on different datasets for retrieval.
Researcher Affiliation Academia National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, Beijing
Pseudocode Yes Algorithm 1 Distributed Hashing
Open Source Code No The paper does not provide any statement or link regarding the public availability of its source code.
Open Datasets Yes CIFAR-101 and GIST-1M2, with footnotes 1http://www.cs.toronto.edu/ kriz/ and 2http://corpus-texmex.irisa.fr/. Also MNIST-8M3 and SIFT10M4. with footnotes 3http://www.csie.ntu.edu.tw/ cjlin/ and 4http://corpus-texmex.irisa.fr/.
Dataset Splits No The paper describes splits for query, database, and training sets, but does not explicitly mention a dedicated validation dataset split for hyperparameter tuning or model selection.
Hardware Specification Yes Each node has a 2.50GHz Intel Xeon CPU.
Software Dependencies No The implementation of our distributed system is based on the Distributed Computing Engine of MATLAB in Linux. (No specific version numbers are provided for MATLAB or other software dependencies).
Experiment Setup Yes Throughout the experiments, we empirically set the penalty parameter ρ as ρ = 100 and the number of ADMM iterations K as K = 5.