Making Convolutional Networks Shift-Invariant Again
Authors: Richard Zhang
ICML 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We test across multiple tasks image classification and image-to-image translation. For Image Net classification, we find, surprisingly, that accuracy increases, indicating effective regularization. Furthermore, we observe better generalization. Performance is more robust and stable to corruptions such as rotation, scaling, blurring, and noise variants. |
| Researcher Affiliation | Industry | 1Adobe Research, San Francisco, CA. Correspondence to: Richard Zhang <rizhang@adobe.com>. |
| Pseudocode | No | The paper does not contain structured pseudocode or algorithm blocks. |
| Open Source Code | Yes | Code and anti-aliased versions of popular networks are available at https://richzhang.github.io/antialiased-cnns/. |
| Open Datasets | Yes | CIFAR Classification To begin, we test classification of low-resolution 32 32 images. The dataset contains 50k training and 10k validation images, classified into one of 10 categories. Image Net Classification We then test on large-scale classification on 224 224 resolution images. The dataset contains 1.2M training and 50k validation images, classified into one of 1000 categories. We test across different architecture families Alex Net (Krizhevsky & Hinton, 2009), VGG (Simonyan & Zisserman, 2015), Res Net (He et al., 2016), Dense Net (Huang et al., 2017), and Mobile Netv2 (Sandler et al., 2018) with different downsampling strategies, as described in Table 1. |
| Dataset Splits | Yes | CIFAR Classification To begin, we test classification of low-resolution 32 32 images. The dataset contains 50k training and 10k validation images, classified into one of 10 categories. Image Net Classification We then test on large-scale classification on 224 224 resolution images. The dataset contains 1.2M training and 50k validation images, classified into one of 1000 categories. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., GPU models, CPU types, memory amounts) used for running its experiments. |
| Software Dependencies | No | The paper mentions "Py Torch (Paszke et al., 2017)" but does not specify a version number for PyTorch or any other software dependencies. |
| Experiment Setup | No | The paper mentions using "standard training schedules from the publicly available Py Torch (Paszke et al., 2017) repository" and describes the types of filters used (Rect-2, Tri-3, Bin-5). However, it does not explicitly state concrete hyperparameter values such as learning rate, batch size, or number of epochs, which are typical for a detailed experimental setup description. |