HiHPQ: Hierarchical Hyperbolic Product Quantization for Unsupervised Image Retrieval

Authors: Zexuan Qiu, Jiahong Liu, Yankai Chen, Irwin King

AAAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on benchmarks show that our proposed method outperforms state-of-the-art baselines.
Researcher Affiliation Academia Zexuan Qiu, Jiahong Liu, Yankai Chen, Irwin King The Chinese University of Hong Kong {zxqiu22, jhliu22, ykchen, king}@cse.cuhk.edu.hk
Pseudocode No The paper does not contain any pseudocode or explicitly labeled algorithm blocks.
Open Source Code No The paper does not include an unambiguous statement about releasing source code for the methodology described, nor does it provide a direct link to a code repository.
Open Datasets Yes The proposed method is evaluated using Flickr25K (Huiskes and Lew 2008), NUS-WIDE (Chua et al. 2009), as well as two experimental protocols CIFAR10 (I) and CIFAR-10 (II) both of which are based on CIFAR-10 (Krizhevsky, Hinton et al. 2009).
Dataset Splits No The paper mentions using CIFAR-10, Flickr25K, and NUS-WIDE datasets but does not explicitly state the specific training, validation, and test splits (percentages, counts, or specific predefined split references) used for the experiments.
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory) used for running its experiments.
Software Dependencies No The paper does not provide specific ancillary software details, such as library names with version numbers, needed to replicate the experiment.
Experiment Setup Yes The pre-defined hierarchies for different datasets are: [50, 20] on Flickr25K; [200, 100, 50] on CIFAR-10 (I); [100, 50, 25] on CIFAR-10 (II); [200, 100, 75] on NUSWIDE. Please refer to the supplementary material for more implementation details.