Semantic bottleneck for computer vision tasks

Maxime Bucher 1, 2 Stéphane Herbin 1 Frédéric Jurie 2
2 Equipe Image - Laboratoire GREYC - UMR6072
GREYC - Groupe de Recherche en Informatique, Image, Automatique et Instrumentation de Caen
Abstract : This paper introduces a novel method for the representation of images that is semantic by nature, addressing the question of computation intelligibility in computer vision tasks. More specifically, our proposition is to introduce what we call a semantic bottleneck in the processing pipeline, which is a crossing point in which the representation of the image is entirely expressed with natural language , while retaining the efficiency of numerical representations. We show that our approach is able to generate semantic representations that give state-of-the-art results on semantic content-based image retrieval and also perform very well on image classification tasks. Intelligibility is evaluated through user centered experiments for failure detection.
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-01909665
Contributor : Maxime Bucher <>
Submitted on : Monday, November 5, 2018 - 4:33:18 PM
Last modification on : Tuesday, April 2, 2019 - 1:36:02 AM
Document(s) archivé(s) le : Wednesday, February 6, 2019 - 2:59:24 PM

Files

main.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01909665, version 1
  • ARXIV : 1811.02234

Citation

Maxime Bucher, Stéphane Herbin, Frédéric Jurie. Semantic bottleneck for computer vision tasks. Asian Conference on Computer Vision (ACCV), Dec 2018, Perth, Australia. ⟨hal-01909665⟩

Share

Metrics

Record views

69

Files downloads

58