Fast Online Hashing with Multi-Label Projection

Authors: Wenzhe Jia, Yuan Cao, Junwei Liu, Jie Gui

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

Reproducibility Variable Result LLM Response
Research Type Experimental The experimental results on two common benchmarks show that the proposed FOH can achieve dramatic superiority on query time up to 6.28 seconds less than state-of-the-art baselines with competitive retrieval accuracy.
Researcher Affiliation Academia 1 Ocean University of China, China 2 State Key Laboratory of Integrated Services Networks (Xidian University), China 3 Southeast University, China 4 Purple Mountain Laboratories, China
Pseudocode No The paper describes algorithms in prose and equations but does not contain any structured pseudocode or algorithm blocks with formal pseudocode syntax.
Open Source Code Yes Our implementation of this paper is publicly available on Git Hub at: https://github.com/caoyuan57/FOH.
Open Datasets Yes In this section, we conduct experiments on two common datasets: CIFAR-10 (Krizhevsky 2009) and FLICKR-25K (Huiskes and Lew 2008) to verify the efficiency and effectiveness of the proposed Fast Online Hashing (FOH).
Dataset Splits No The paper provides detailed information on training and test set creation and partitioning into blocks for stream data, but does not explicitly mention a separate 'validation' dataset split.
Hardware Specification No The paper mentions "the Big Data Computing Center of Southeast University for providing the facility support on the numerical calculations" but does not provide specific hardware details such as CPU/GPU models or memory specifications.
Software Dependencies No The paper does not provide specific ancillary software details with version numbers (e.g., library or solver names with version numbers like Python 3.8, CPLEX 12.4).
Experiment Setup Yes Here, we provide the exact values of the parameter configurations in Tab. ??. Review that u denotes the number of the central points in the query pool, v denotes the number of the nearest neighbors of each central point, β denotes the number of the returned central points when a new query arrives, {σ, θ, µ, λ, τ} denotes the hyper-parameters in the objective function. Table 2: Parameter configurations on two datasets.