Decoupled Box Proposal and Featurization with Ultrafine-Grained Semantic Labels Improve Image Captioning and Visual Question Answering

@article{Changpinyo2019DecoupledBP,
  title={Decoupled Box Proposal and Featurization with Ultrafine-Grained Semantic Labels Improve Image Captioning and Visual Question Answering},
  author={Soravit Changpinyo and Bo Pang and Piyush Sharma and Radu Soricut},
  journal={ArXiv},
  year={2019},
  volume={abs/1909.02097}
}
  • Soravit Changpinyo, Bo Pang, +1 author Radu Soricut
  • Published 2019
  • Computer Science
  • ArXiv
  • Object detection plays an important role in current solutions to vision and language tasks like image captioning and visual question answering. [...] Key Result Empirically, we demonstrate that this leads to effective transfer learning and improved image captioning and visual question answering models, as measured on publicly available benchmarks.Expand Abstract
    5 Citations

    Figures, Tables, and Topics from this paper.

    References

    SHOWING 1-10 OF 45 REFERENCES
    Learning to Count Objects in Natural Images for Visual Question Answering
    • 101
    • PDF
    Exploring Models and Data for Image Question Answering
    • 439
    • Highly Influential
    • PDF
    Visual7W: Grounded Question Answering in Images
    • 432
    • PDF
    Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering
    • 631
    • Highly Influential
    • PDF
    Visual and Semantic Knowledge Transfer for Large Scale Semi-Supervised Object Detection
    • 24
    • PDF
    Captioning Images with Diverse Objects
    • 92
    • PDF
    Bottom-Up and Top-Down Attention for Image Captioning and Visual Question Answering
    • 1,129
    • Highly Influential
    • PDF
    Large Scale Visual Recognition through Adaptation using Joint Representation and Multiple Instance Learning
    • 14
    • PDF
    GQA: a new dataset for compositional question answering over real-world images
    • 51
    • PDF
    Areas of Attention for Image Captioning
    • 106
    • PDF