Raven’s Progressive Matrices Completion with Latent Gaussian Process Priors
Authors: Fan Shi, Bin Li, Xiangyang Xue9612-9620
AAAI 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We evaluate the proposed model on RPM-like datasets with multiple continuously-changing visual concepts. Experimental results demonstrate that our model requires only few training samples to paint high-quality answers, generate novel RPM panels, and achieve interpretability through concept-specific latent variables.We evaluate the model in RPM-like datasets with continuously changing visual concepts. To evaluate both the reconstruction quality and comprehension of underlying conceptchanging rules, we calculate the averaged MSE over panel cells by leaving out and predicting one cell in turn. For all datasets, the proposed model outperforms the VAE-GAN inpainting model (Yu et al. 2018). |
| Researcher Affiliation | Academia | Fan Shi, Bin Li*, Xiangyang Xue Shanghai Key Laboratory of Intelligent Information Processing School of Computer Science, Fudan University {fshi19, libin, xyxue}@fudan.edu.cn |
| Pseudocode | No | The paper describes the generative and inference models with diagrams and textual explanations, but it does not include formal pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide any statement or link indicating that the source code for the described methodology is publicly available. |
| Open Datasets | No | We create new RPM-like datasets that remove selection panels and retain complete 9-cell panels.Results about more instances of the RPM-like dataset and the selective task accuracy compared with RPM solvers are exhibited in the supplementary material.The paper describes creating its own datasets ("Polygon", "Circle") but does not provide any specific link, DOI, or formal citation for public access to these datasets. |
| Dataset Splits | Yes | Each instance of the dataset contains a disjoint training set, validation set, and test set. For further performance evaluation, we provide various-sized (from 50 to 50000) training sets and one fixed 10000-sample test set. |
| Hardware Specification | No | The paper does not specify the hardware used for running the experiments. |
| Software Dependencies | No | The paper mentions various models and frameworks (e.g., VAE, GP) but does not list specific software dependencies with version numbers. |
| Experiment Setup | No | The paper mentions hyperparameters β and γ, but it does not provide their specific values or other detailed experimental setup configurations like learning rate, batch size, or optimizer settings. |