Beyond RPCA: Flattening Complex Noise in the Frequency Domain
Authors: Yunhe Wang, Chang Xu, Chao Xu, Dacheng Tao
AAAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on synthetic data and video background subtraction demonstrate that FRPCA is effective for handles complex noise. Experiments We next evaluated the performance of the proposed method and compared it with the state-of-the-art methods on synthetic data and real video data. |
| Researcher Affiliation | Academia | 1Key Laboratory of Machine Perception (MOE), School of EECS, Peking University, China 2Centre for Artificial Intelligence, University of Technology Sydney, Australia 3Cooperative Medianet Innovation Center, Peking University, China |
| Pseudocode | Yes | Algorithm 1 Frequency RPCA for data recovery. Input: Observed data Y, weight parameter λ; 1: Convert Y into the frequency domain CY SY;... |
| Open Source Code | No | The paper does not provide concrete access to its own source code (e.g., a specific repository link or an explicit code release statement). |
| Open Datasets | Yes | Four commonly utilized video sequences were selected for this experiment: Bootstrap, Hall, Fountain, and Campus (Li et al. 2004), containing 3055, 3584, 523, and 1439 frames, respectively. Five famous and commonly used pictures were selected as examples: Peppers, Lena, Butterfly, Baboon, and Cameraman. |
| Dataset Splits | No | The paper describes dataset generation and the extraction of frames for experiments but does not provide specific training/validation/test dataset splits (e.g., percentages, sample counts, or citations to predefined splits) needed for reproducibility. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., exact GPU/CPU models, memory amounts, or detailed computer specifications) used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific ancillary software details (e.g., library or solver names with version numbers) needed to replicate the experiment. |
| Experiment Setup | Yes | The proposed FRPCA only has one hyper parameter λ, which was empirically set as 10 m 1/2 for the synthetic data and m 1/2 for the video data, where m is the dimensionality of each sample. We set rank(X) = 40 for these methods, and tuned their parameters to make results comparable. |