Fast Structural Binary Coding
Authors: Dongjin Song, Wei Liu, David A. Meyer
IJCAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Empirical studies based upon two image datasets demonstrate that the proposed binary coding approaches achieve superior image search performance to the states-of-the-art. |
| Researcher Affiliation | Collaboration | Dongjin Song , Wei Liu], and David A. Meyer Department of Electrical and Computer Engineering,University of California, San Diego La Jolla, USA, 92093-0409. Email: dosong@ucsd.edu ] Didi Research, Didi Kuaidi, Beijing, China. Email: wliu@ee.columbia.edu Department of Mathematics,University of California, San Diego La Jolla, USA, 92093-0112. Email: dmeyer@math.ucsd.edu |
| Pseudocode | Yes | Algorithm 1 Fast Structural Binary Coding |
| Open Source Code | No | The paper states 'We implement the proposed FSBC and baseline algorithms using Matlab' but does not provide a link or explicit statement about open-sourcing the code. |
| Open Datasets | Yes | In SUN397, 100 images are randomly sampled from each of the 18 largest scene categories to form a test set of 1,800 query images... In You Tube Faces, 100 face images from each of the 65 largest face classes are randomly sampled to form a test set of 6,500 query images. |
| Dataset Splits | Yes | For supervised methods, we randomly choose 200 images from each of the 18 scene categories to form a training set of 3,600 images; an additional 50 images from each of these 18 scene categories are randomly selected to form a validation set of 900 query images. All the rest of the images in the 397 categories are then used as the database samples. For supervised learning, 1,000 images from each of the 65 face classes are randomly draw to form a training set of 65,000 face images; an additional 50 images from each of these 65 scene categories are randomly selected to form a validation set of 3250 query images. |
| Hardware Specification | Yes | We implement the proposed FSBC and baseline algorithms using Matlab on a PC with Intel Core i7-4770K Processor 3.5GHz and 32GB RAM. |
| Software Dependencies | No | The paper mentions 'Matlab' but does not provide specific version numbers or other software dependencies with version information. |
| Experiment Setup | Yes | The parameters λ and µ of FSBC are determined by cross validation over the grid {1, 10 1, 10 2, 10 3, 10 4, 10 5, 10 6}. We will discuss the parameter sensitivity later. ... is set to be 1 in all our experiments. |