Robust Multi-View Subspace Learning through Dual Low-Rank Decompositions

Authors: Zhengming Ding, Yun Fu

AAAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on two multi-view benchmarks, e.g., face and object images, have witnessed the superiority of our proposed algorithm, by comparing it with the state-of-the-art algorithms.
Researcher Affiliation Academia Department of Electrical & Computer Engineering, Northeastern University, Boston, USA College of Computer & Information Science, Northeastern University, Boston, USA allanding@ece.neu.edu, yunfu@ece.neu.edu
Pseudocode Yes Algorithm 1 Solution to Problem (3)
Open Source Code No The paper does not provide an explicit statement or link for the open-source code of their methodology.
Open Datasets Yes CMU-PIE Face database is consisted of 68 subjects in total, which is a multi-view face dataset 1 and show large variances within the same subject but in different poses. ... 1http://vasc.ri.cmu.edu/idb/html/face/. COIL-100 object database2 contains 100 categories with 7200 images. ... 2http://www.cs.columbia.edu/CAVE/software/softlib/coil100.php
Dataset Splits No The paper describes training and testing splits, but does not explicitly mention a separate validation set or provide details for a validation split.
Hardware Specification Yes Experiments are conducted with Matlab 2014b, CPU i7-3770 and 32 GB memory size.
Software Dependencies Yes Experiments are conducted with Matlab 2014b, CPU i7-3770 and 32 GB memory size.
Experiment Setup Yes In Algorithm 1, where we set those parameters μ0, ρ, ϵ, tmax and μmax empirically, while tuning the two trade-offs, i.e., λ and α throughout the experiment, which is further discussed in experimental part. Therefore, we set λ = 10 2 and α = 102 throughout the experiments.