Supervised Hashing via Uncorrelated Component Analysis
Authors: SungRyull Sohn, Hyunwoo Kim, Junmo Kim
AAAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We performed image retrieval experiments on three image benchmarks: Yahoo! (Thomee et al. 2013), Trevi and Halfdome photo-tourism (Snavely, Seitz, and Szeliski 2006). ... The results show that UCA-based hashing outperforms state-of-the-art methods, and has computationally efficient training and encoding processes. |
| Researcher Affiliation | Collaboration | Sung Ryull Sohn CG Research Team Electronics and Telecommunications Research Institute School of Electrical Engineering Korea Advanced Institute of Science and Technology sungluol@etri.re.kr Hyunwoo Kim Kakao Corp. eugene.kim@kakaocorp.com Junmo Kim School of Electrical Engineering Korea Advanced Institute of Science and Technology junmo@ee.kaist.ac.kr |
| Pseudocode | No | The paper does not contain any sections or figures explicitly labeled 'Pseudocode' or 'Algorithm', nor are there structured, code-like steps for any method. |
| Open Source Code | No | The paper does not provide any statement or link indicating that the source code for their proposed UCAH method is publicly available. |
| Open Datasets | Yes | We performed image retrieval experiments on three image benchmarks: Yahoo! (Thomee et al. 2013), Trevi and Halfdome photo-tourism (Snavely, Seitz, and Szeliski 2006). ... We also performed the scalability test, by adding 1M distractor images from the Flickr-1M (Huiskes and Lew 2008) dataset to the database of Yahoo! as the false positive samples. |
| Dataset Splits | Yes | The images of each dataset were divided into query and database. 1000 images were chosen at random for the query, and the rest formed the database. 5K images in the database were randomly sampled and used for training. |
| Hardware Specification | Yes | All the experiments were implemented in MATLAB and performed on a desktop with an Intel Core i7 CPU at 3.30 GHz with 32 GB RAM. |
| Software Dependencies | No | The paper states 'All the experiments were implemented in MATLAB' but does not specify a version number for MATLAB or any other software dependencies. |
| Experiment Setup | No | The paper mentions data splits and feature descriptors (GIST 320-dimensional, 384-dimensional) and code length 'k', but does not provide specific hyperparameters for model training such as learning rates, optimizers, batch sizes, or number of epochs. |