Learning Robust Multi-view Representation Using Dual-masked VAEs
Authors: Jiedong Wang, Kai Guo, Peng Hu, Xi Peng, Hao Wang
IJCAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on four multi-view datasets show the effectiveness of the proposed method and its superior performance in comparison to baselines. We conduct experiments on four publicly available multi-view datasets. The statistics of each dataset are shown in Table 1. |
| Researcher Affiliation | Academia | 1College of Computer Science, Sichuan University, China 2National Key Laboratory of Fundamental Algorithms and Models for Engineering Numerical Simulation, Sichuan University, China |
| Pseudocode | No | The paper describes the methodology using textual explanations and mathematical formulations, but it does not contain any explicitly labeled pseudocode or algorithm blocks. For example, Section 3 'Methodology' and its subsections describe the process in detail but without a structured algorithm format. |
| Open Source Code | Yes | The code is available at https://github.com/XLearning-SCU/ 2025-IJCAI-Dual VAE. |
| Open Datasets | Yes | To evaluate the proposed model, we conduct experiments on four publicly available multi-view datasets. The statistics of each dataset are shown in Table 1. COIL-20 [Nene et al., 1996b] COIL-100 [Nene et al., 1996a] E-MNIST [Liu and Tuzel, 2016] Poly MNIST [Palumbo et al., 2023] |
| Dataset Splits | No | The paper defines View-Missing Ratio (VMR) and Sample-Noise Ratio (SNR) for evaluating the model's robustness under different conditions, and states that these ratios are varied from 0% to 90%. However, it does not provide specific training/validation/test splits (e.g., 80/10/10 split percentages or sample counts) for the model's primary training process. The experimental results are based on clustering and classification tasks on 'learnt representations', implying a training phase, but the split details for this phase are not explicitly provided. |
| Hardware Specification | Yes | We deployed the proposed model and baselines using Pytorch 2.3.1 and ran experiments on NVIDIA TITAN GPUs with 24GB of memory in Ubuntu 20.04.1 LTS. |
| Software Dependencies | Yes | We deployed the proposed model and baselines using Pytorch 2.3.1 and ran experiments on NVIDIA TITAN GPUs with 24GB of memory in Ubuntu 20.04.1 LTS. |
| Experiment Setup | Yes | We utilize Adam optimizer with weight decay and set training epochs as 200. Moreover, we set the initial learning rate to 1e-4 and adopt cosine annealing during the training. Following [He et al., 2022; Ke et al., 2024], the default patch-mask ratio is 0.7. We then initially set view-mask ratio as 0.3. Both the dimensions of view-consistent representation and view-specific representation are 10. |