Nonnegative Orthogonal Graph Matching
Authors: Bo Jiang, Jin Tang, Chris Ding, Bin Luo
AAAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Promising experimental results demonstrate benefits of NOGM algorithm. |
| Researcher Affiliation | Academia | 1School of Computer Science and Technology, Anhui University, Hefei, 230601, China 2CSE Department, University of Texas at Arlington, Arlington, TX 76019, USA |
| Pseudocode | No | The paper describes the update rule as mathematical equations, but does not present it as a structured pseudocode or algorithm block. |
| Open Source Code | No | The paper does not provide any explicit statements about open-sourcing code, nor does it include links to a code repository. |
| Open Datasets | Yes | We perform feature matching on CMU and YORK image sequences (Cho, Lee, and Lee 2010; Caetano et al. 2009; Luo, Wilson, and Hancock 2003). ... Our first evaluation is conducted on the image pairs selected from Caltech-101 and MSRC datasets (Cho, Lee, and Lee 2010). ... Our second evaluation is performed on the image pairs (20 pairs) selected from Zurich Building Image Database (Zu Bud) (Ng and Kingsbury 2010). |
| Dataset Splits | No | The paper mentions generating random graph pairs and matching images but does not specify explicit training, validation, or test dataset splits (e.g., percentages, counts, or specific split files) for reproducibility. |
| Hardware Specification | No | The paper does not provide any specific hardware details (e.g., CPU/GPU models, memory) used for running the experiments. |
| Software Dependencies | No | The paper does not specify any software dependencies with version numbers. |
| Experiment Setup | Yes | Following the experimental setting (Cho, Lee, and Lee 2010), we first generated two graphs GM and GD. ... The corresponding edge r D ij in GD was perturbed by adding a random Gaussian noise σ to r M ij . Here, σ controls the level of deformation noise. The affinity matrix W is computed by Wij,kl = exp( r D ik r M jl 2 F σr ), where scaling factor σr has been set to 0.025 in this experiment. For each noise level σ or nout, we have generated 100 random graph pairs and then computed the average performances including matching accuracy, objective score, sparsity and orthogonality. |