Deep Diffusion-Invariant Wasserstein Distributional Classification

Authors: Sung Woo Park, Dong Wook Shu, Junseok Kwon

NeurIPS 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental For empirical validation, we applied our method to two classification problems: 3D point cloud classification using the Model Net10 (28) dataset and image classification using the CIFAR10 dataset (12), where the data suffer from various perturbations. ... Table 1: 2D image classification accuracy (in %) on the CIFAR10 dataset with perturbations. ... Table 2: 3D point cloud classification accuracy (in %) for the Model Net10 dataset with perturbations.
Researcher Affiliation Academia Sung Woo Park Dong Wook Shu Junseok Kwon School of Computer Science and Engineering Chung-Ang University, Seoul, Korea pswkiki@gmail.com seowok@naver.com jskwon@cau.ac.kr
Pseudocode Yes Algorithm 1 Deep WDC
Open Source Code No The paper does not provide any explicit statement or link indicating that the source code for the described methodology is publicly available.
Open Datasets Yes For empirical validation, we applied our method to two classification problems: 3D point cloud classification using the Model Net10 (28) dataset and image classification using the CIFAR10 dataset (12)
Dataset Splits No The paper mentions using CIFAR10 and ModelNet10 datasets but does not explicitly provide specific details about the training, validation, or test splits (e.g., percentages or sample counts).
Hardware Specification Yes All experiments were executed using a single RTX 2080 TI GPU.
Software Dependencies Yes Our method was implemented using Pytorch1.4.0 and Python3.6.
Experiment Setup Yes We used the ADAM optimizer with a learning rate of 10 5 for the network g and a learning rate of 10 3 for the network f, as well as for the baseline networks.