Universe Points Representation Learning for Partial Multi-Graph Matching

Authors: Zhakshylyk Nurlanov, Frank R. Schmidt, Florian Bernard

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

Reproducibility Variable Result LLM Response
Research Type Experimental The proposed approach advances the state of the art in semantic keypoint matching problem, evaluated on Pascal VOC, CUB, and Willow datasets. Moreover, the set of controlled experiments on a synthetic graph matching dataset demonstrates the scalability of our method to graphs with large number of nodes and its robustness to high partiality.
Researcher Affiliation Collaboration Zhakshylyk Nurlanov1,2, Frank R. Schmidt1, Florian Bernard2 1 Bosch Center for Artificial Intelligence 2 University of Bonn
Pseudocode No The paper does not contain any explicit pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any explicit statement about releasing source code or a link to a code repository for the described methodology.
Open Datasets Yes We evaluate the real-world performance on Pascal VOC and CUB-200-2011 datasets. We also estimate the transferability of the learned node feature encoder to other datasets (Pascal Willow).
Dataset Splits No The paper mentions using 'training set' and 'test set' but does not provide specific details on the dataset splits (e.g., percentages, sample counts, or explicit citations for the splits themselves) needed for reproduction.
Hardware Specification No The paper mentions 'Training of BBGM-Multi-3 on problems with more than 300 universe points does not fit into GPU memory (48 GB)'. While a memory capacity is given, no specific GPU model or type is mentioned.
Software Dependencies No The paper mentions software components like 'Spline CNN' and references 'Image Net pre-trained VGG features' but does not specify any version numbers for these or other software dependencies.
Experiment Setup No The paper describes the architecture and components used (e.g., Spline CNN blocks, MLPz) but does not provide specific experimental setup details such as hyperparameter values (learning rate, batch size, epochs, optimizer settings) or training configurations.