Discovery of Single Independent Latent Variable
Authors: Uri Shaham, Jonathan Svirsky, Ori Katz, Ronen Talmon
NeurIPS 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We demonstrate the performance of the proposed approach in several tasks, including image synthesis, voice cloning, and fetal ECG extraction. Experimentally, we show both analysis and synthesis results on several datasets, consisting of simulated and real-world data. |
| Researcher Affiliation | Academia | Uri Shaham Department of Computer Science Bar-Ilan University Ramat Gan, Israel uri.shaham@biu.ac.il Jonathan Svirsky Faculty of Engineering Bar-Ilan University Ramat Gan, Israel svirskj@biu.ac.il Ori Katz Electrical and Computer Engineering Technion Haifa, Israel orikats@campus.technion.ac.il Ronen Talmon Electrical and Computer Engineering Technion Haifa, Israel ronen@ee.technion.ac.il |
| Pseudocode | No | The paper describes the model architecture and training objectives in text and equations, but does not include any explicitly labeled "Pseudocode" or "Algorithm" block. |
| Open Source Code | Yes | In addition, codes reproducing some of the results in this manuscript are available at https://github.com/shaham-lab/disilv. |
| Open Datasets | Yes | We use a subset of the corpus containing all the utterances for the first 30 speakers (p225p256, without p235 and p242). We consider the dataset from [52], which is publicly available3 on Physio Net [16]. |
| Dataset Splits | No | No explicit train/validation/test splits given in percentages or counts. The paper mentions evaluating on a "subset of the data containing parallel sentences" for MCD, but does not detail how this subset is defined relative to the training data nor does it refer to it as a validation set. For ECG, it states training a model per subject on "input-condition pairs", but no split information is provided. |
| Hardware Specification | No | The paper does not provide specific details about the hardware used for running the experiments, such as GPU models, CPU specifications, or memory. |
| Software Dependencies | No | The paper mentions software components like "melgan [34] vocoder" and "Jasper [38] blocks" and a "script provided in [37]" but does not specify any version numbers for these or other software dependencies. |
| Experiment Setup | Yes | The network architectures and training hyperparameters used in each of the experiments are described Appendix ??. |