{
  "id": "1511.03042",
  "version": "v1",
  "published": "2015-11-10T09:54:20.000Z",
  "updated": "2015-11-10T09:54:20.000Z",
  "title": "Analyzing Stability of Convolutional Neural Networks in the Frequency Domain",
  "authors": [
    "Elnaz J. Heravi",
    "Hamed H. Aghdam",
    "Domenec Puig"
  ],
  "comment": "Under review as a conference paper at ICLR2016",
  "categories": [
    "cs.CV"
  ],
  "abstract": "Understanding the internal process of ConvNets is commonly done using visualization techniques. However, these techniques do not usually provide a tool for estimating the stability of a ConvNet against noise. In this paper, we show how to analyze a ConvNet in the frequency domain using a 4-dimensional visualization technique. Using the frequency domain analysis, we show the reason that a ConvNet might be sensitive to a very low magnitude additive noise. Our experiments on a few ConvNets trained on different datasets revealed that convolution kernels of a trained ConvNet usually pass most of the frequencies and they are not able to effectively eliminate the effect of high frequencies. Our next experiments shows that a convolution kernel which has a more concentrated frequency response could be more stable. Finally, we show that fine-tuning a ConvNet using a training set augmented with noisy images can produce more stable ConvNets.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2015-11-10T09:54:20.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "convolutional neural networks",
      "analyzing stability",
      "convolution kernel",
      "visualization technique",
      "frequency domain analysis"
    ],
    "tags": [
      "conference paper"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable",
      "adsabs": "2015arXiv151103042H"
    }
  }
}