Conference paper Open Access

INPUT COMPLEXITY AND OUT-OF-DISTRIBUTION DETECTION WITH LIKELIHOOD-BASED GENERATIVE MODELS

Serra, Joan; Álvarez, David; Gómez, Vicenç; Slizovskaia , Olga; F. Núñez, José; Luque, Jordi


JSON-LD (schema.org) Export

{
  "description": "<p>Likelihood-based generative models are a promising resource to detect out-of- distribution (OOD) inputs which could compromise the&nbsp;robustness or reliability of a machine learning system. However, likelihoods derived from such models have been shown to be problematic for detecting certain types of inputs that sig- nificantly differ from training data. In this paper, we pose that this problem is due to the excessive influence that input complexity has in generative models&rsquo; likelihoods. We report a set of experiments supporting this hypothesis, and use an estimate of input complexity to derive an efficient and parameter-free OOD score, which can be seen as a likelihood-ratio, akin to Bayesian model compari- son. We find such score to perform comparably to, or even better than, existing OOD detection approaches under a wide range of data sets, models, model sizes, and complexity estimates.</p>", 
  "license": "https://creativecommons.org/licenses/by/4.0/legalcode", 
  "creator": [
    {
      "affiliation": "Dolby Laboratories,  Barcelona, Spain", 
      "@type": "Person", 
      "name": "Serra, Joan"
    }, 
    {
      "affiliation": "Universitat Polit\u00e8cnica de Catalunya,  Barcelona, Spain", 
      "@type": "Person", 
      "name": "\u00c1lvarez, David"
    }, 
    {
      "affiliation": "Universitat Pompeu Fabra, Barcelona, Spain", 
      "@type": "Person", 
      "name": "G\u00f3mez, Vicen\u00e7"
    }, 
    {
      "affiliation": "Universitat Pompeu Fabra, Barcelona, Spain", 
      "@type": "Person", 
      "name": "Slizovskaia , Olga"
    }, 
    {
      "affiliation": "Universitat Pompeu Fabra, Barcelona, Spain", 
      "@type": "Person", 
      "name": "F. N\u00fa\u00f1ez, Jos\u00e9"
    }, 
    {
      "affiliation": "Telef\u00f3nica I+D, Research, Barcelona, Spain", 
      "@type": "Person", 
      "name": "Luque, Jordi"
    }
  ], 
  "headline": "INPUT COMPLEXITY AND OUT-OF-DISTRIBUTION DETECTION WITH LIKELIHOOD-BASED GENERATIVE MODELS", 
  "image": "https://zenodo.org/static/img/logos/zenodo-gradient-round.svg", 
  "datePublished": "2020-04-30", 
  "url": "https://zenodo.org/record/5253740", 
  "@type": "ScholarlyArticle", 
  "keywords": [], 
  "@context": "https://schema.org/", 
  "identifier": "https://doi.org/10.5281/zenodo.5253740", 
  "@id": "https://doi.org/10.5281/zenodo.5253740", 
  "workFeatured": {
    "url": "https://iclr.cc", 
    "alternateName": "ICLR 2020", 
    "location": "Addiss Abbeba", 
    "@type": "Event", 
    "name": "Ninth International Conference on Learning Representation"
  }, 
  "name": "INPUT COMPLEXITY AND OUT-OF-DISTRIBUTION DETECTION WITH LIKELIHOOD-BASED GENERATIVE MODELS"
}
17
12
views
downloads
All versions This version
Views 1717
Downloads 1212
Data volume 23.3 MB23.3 MB
Unique views 1414
Unique downloads 1212

Share

Cite as