ProMix: Combating Label Noise via Maximizing Clean Sample Utility
Authors: Ruixuan Xiao, Yiwen Dong, Haobo Wang, Lei Feng, Runze Wu, Gang Chen, Junbo Zhao
IJCAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this section, we present the main results and part of the ablation results to demonstrate the effectiveness of the Pro Mix framework. We put more experimental results, including results on imbalanced dataset with synthetic noise, results on dataset with instance-dependent label noise and more results for ablation experiment in Appendix B. |
| Researcher Affiliation | Collaboration | Ruixuan Xiao1 , Yiwen Dong1 , Haobo Wang1 , Lei Feng2 , Runze Wu3 , Gang Chen1 and Junbo Zhao1 1Zhejiang University, Hangzhou, China 2Nanyang Technological University, Singapore 3Net Ease Fuxi AI Lab, Hangzhou, China |
| Pseudocode | No | The paper contains diagrams illustrating the framework and selection process, but no formal pseudocode or algorithm blocks with numbered steps or code-like formatting. |
| Open Source Code | No | The paper does not contain an explicit statement or link indicating that source code for the described methodology is publicly available. |
| Open Datasets | Yes | We first evaluate the performance of Pro Mix on CIFAR-10, CIFAR-100 [Krizhevsky et al., 2009], Clothing1M [Xiao et al., 2015] and ANIMAL-10N [Song et al., 2019] dataset. For CIFAR-10/100, we conduct experiments with synthetic symmetric and asymmetric label noise following the previous protocol [Tanaka et al., 2018]. |
| Dataset Splits | No | The paper mentions using datasets like CIFAR-10 and CIFAR-100 which have standard splits, but it does not explicitly state the percentages or sample counts for training, validation, and test splits within the paper's text. |
| Hardware Specification | No | The paper mentions using 'Res Net-18 as the backbone' but does not specify any hardware details such as GPU models, CPU types, or memory used for running the experiments. |
| Software Dependencies | No | The paper mentions software like SGD and Rand Augment, but does not provide specific version numbers for any software, libraries, or frameworks used in the experiments. |
| Experiment Setup | Yes | For CIFAR experiments, we use two Res Net-18 as the backbone of the peer networks. These networks are trained for 600 epochs, with a warm-up period of 10 epochs for CIFAR-10 and 30 epochs for CIFAR-100. We employ SGD as the optimizer with a momentum of 0.9 and weight decay of 5e 4. The batch size is fixed as 256 and the initial learning rate is 0.05, which decays by a cosine scheduler. The threshold τ is set as 0.99 and 0.95 for CIFAR-10 and CIFAR-100. The debiasing factor α is set as 0.8 and 0.5 for CIFAR-10 and CIFAR-100. m is fixed as 0.9999. We leverage Rand Augment to generate a strongly augmented view for consistency training. |