Corpus ID: 551912

Visalogy: Answering Visual Analogy Questions

  title={Visalogy: Answering Visual Analogy Questions},
  author={F. Sadeghi and C. L. Zitnick and Ali Farhadi},
  • F. Sadeghi, C. L. Zitnick, Ali Farhadi
  • Published in NIPS 2015
  • Computer Science
  • In this paper, we study the problem of answering visual analogy questions. These questions take the form of image A is to image B as image C is to what. Answering these questions entails discovering the mapping from image A to image B and then extending the mapping to image C and searching for the image D such that the relation from A to B holds for C to D. We pose this problem as learning an embedding that encourages pairs of analogous images with similar transformations to be close together… CONTINUE READING

    Figures and Topics from this paper.

    Visual Dynamics: Probabilistic Future Frame Synthesis via Cross Convolutional Networks
    • 316
    • Open Access
    Modeling Context in Referring Expressions
    • 202
    • Open Access
    Deep Feature Interpolation for Image Content Changes
    • 133
    • Open Access
    Neural Scene De-rendering
    • 77
    • Open Access
    From Red Wine to Red Tomato: Composition with Context
    • 74
    • Open Access
    Deep Manifold Traversal: Changing Labels with Convolutional Features
    • 49
    • Open Access
    Detecting Unseen Visual Relations Using Analogies
    • 24
    • Open Access
    From A to Z: Supervised Transfer of Style and Content Using Deep Neural Network Generators
    • 25
    • Open Access
    Quartet-net Learning for Visual Instance Retrieval
    • 9


    Publications referenced by this paper.
    ImageNet Classification with Deep Convolutional Neural Networks
    • 51,777
    • Open Access
    Caffe: Convolutional Architecture for Fast Feature Embedding
    • 12,062
    • Open Access
    VQA: Visual Question Answering
    • 1,823
    • Open Access
    Linguistic Regularities in Continuous Space Word Representations
    • 2,486
    • Open Access
    Image analogies
    • 1,474
    • Open Access