A-Fast-RCNN: Hard Positive Generation via Adversary for Object Detection

@article{Wang2017AFastRCNNHP,
  title={A-Fast-RCNN: Hard Positive Generation via Adversary for Object Detection},
  author={X. Wang and Abhinav Shrivastava and A. Gupta},
  journal={2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
  year={2017},
  pages={3039-3048}
}
  • X. Wang, Abhinav Shrivastava, A. Gupta
  • Published 2017
  • Computer Science
  • 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • How do we learn an object detector that is invariant to occlusions and deformations. [...] Key Method In our framework both the original detector and adversary are learned in a joint manner. Our experimental results indicate a 2.3% mAP boost on VOC07 and a 2.6% mAP boost on VOC2012 object detection challenge compared to the Fast-RCNN pipeline.Expand Abstract
    305 Citations

    Figures, Tables, and Topics from this paper

    Occlusion Problem-Oriented Adversarial Faster-RCNN Scheme
    • 1
    • Highly Influenced
    • PDF
    Generative Adversarial Learning Towards Fast Weakly Supervised Detection
    • 43
    • PDF
    Towards Automated Testing and Robustification by Semantic Adversarial Data Generation
    • 1
    • PDF
    Occlusion Handling in Generic Object Detection: A Review
    • PDF
    SINT++: Robust Visual Tracking via Adversarial Positive Instance Generation
    • 61
    • PDF
    VITAL: VIsual Tracking via Adversarial Learning
    • 232
    • PDF
    Adversarial Zero-shot Learning With Semantic Augmentation
    • 14
    MSANet: multimodal self-augmentation and adversarial network for RGB-D object recognition
    • 8

    References

    SHOWING 1-10 OF 60 REFERENCES
    R-FCN: Object Detection via Region-based Fully Convolutional Networks
    • 2,895
    • PDF
    Learning to Segment Object Candidates
    • 553
    • PDF
    Improved Techniques for Training GANs
    • 4,055
    • PDF
    Training Region-Based Object Detectors with Online Hard Example Mining
    • 1,152
    • PDF
    Context Encoders: Feature Learning by Inpainting
    • 2,284
    • PDF
    Unsupervised and Semi-supervised Learning with Categorical Generative Adversarial Networks
    • 478
    • PDF
    You Only Look Once: Unified, Real-Time Object Detection
    • 10,566
    • PDF
    SSD: Single Shot MultiBox Detector
    • 9,897
    • PDF
    Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
    • 21,738
    • PDF
    DenseBox: Unifying Landmark Localization with End to End Object Detection
    • 219
    • PDF