A Bayesian Data Augmentation Approach for Learning Deep Models
Authors: Toan Tran, Trung Pham, Gustavo Carneiro, Lyle Palmer, Ian Reid
NeurIPS 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Classification results on MNIST, CIFAR-10 and CIFAR-100 show the better performance of our proposed method compared to the current dominant data augmentation approach mentioned above the results also show that our approach produces better classification results than similar GAN models. |
| Researcher Affiliation | Academia | Toan Tran1, Trung Pham1, Gustavo Carneiro1, Lyle Palmer2 and Ian Reid1 1School of Computer Science, 2School of Public Health The University of Adelaide, Australia {toan.m.tran, trung.pham, gustavo.carneiro, lyle.palmer, ian.reid} @adelaide.edu.au |
| Pseudocode | No | The paper does not contain structured pseudocode or algorithm blocks. |
| Open Source Code | Yes | In this section, we compare our proposed Bayesian DA algorithm with the commonly used DA technique [19] (denoted as PMDA) on several image classification tasks (code available at: https: //github.com/toantm/keras-bda). |
| Open Datasets | Yes | This comparison is based on experiments using the following three datasets: MNIST [20] (containing 60, 000 training and 10, 000 testing images of 10 handwritten digits), CIFAR-10[18] (consisting of 50, 000 training and 10, 000 testing images of 10 visual classes like car, dog, cat, etc.), and CIFAR-100 [18] (containing the same amount of training and testing samples as CIFAR-10, but with 100 visual classes). |
| Dataset Splits | No | The paper mentions training and testing sets for each dataset (e.g., '60,000 training and 10,000 testing images'), but does not explicitly provide details about a separate validation split. |
| Hardware Specification | No | The paper does not provide specific hardware details (like GPU/CPU models or memory specifications) used for running the experiments. |
| Software Dependencies | No | The paper mentions using Keras (implied by the GitHub repository name 'keras-bda') and specific optimizers (Adadelta, Adam, SGD), but no version numbers for any software dependencies are provided. |
| Experiment Setup | Yes | For training, we use Adadelta (with learning rate=1.0, decay rate=0.95 and epsilon=1e 8) for the Classifier (C), Adam (with learning rate 0.0002, and exponential decay rate 0.5) for the Generator (G) and SDG (with learning rate 0.01) for the Authenticator (A). The noise vector used by the Generator G is based on a standard Gaussian noise. In all experiments, we use training batches of size 100. |