NeuralNetwork-Viterbi: A Framework for Weakly Supervised Video Learning

@article{Richard2018NeuralNetworkViterbiAF,
  title={NeuralNetwork-Viterbi: A Framework for Weakly Supervised Video Learning},
  author={A. Richard and Hilde Kuehne and Ahsan Iqbal and Juergen Gall},
  journal={2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year={2018},
  pages={7386-7395}
}
  • A. Richard, Hilde Kuehne, +1 author Juergen Gall
  • Published 2018
  • Computer Science
  • 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition
  • Video learning is an important task in computer vision and has experienced increasing interest over the recent years. Since even a small amount of videos easily comprises several million frames, methods that do not rely on a frame-level annotation are of special importance. In this work, we propose a novel learning algorithm with a Viterbi-based loss that allows for online and incremental learning of weakly annotated video data. We moreover show that explicit context and length modeling leads… CONTINUE READING
    37 Citations

    Figures, Tables, and Topics from this paper.

    Learning Video Representations using Contrastive Bidirectional Transformer
    • 59
    • Highly Influenced
    Few-Shot Video Classification via Temporal Alignment
    • 14
    • PDF
    SCT: Set Constrained Temporal Transformer for Set Supervised Action Segmentation
    • M. Fayyaz, Juergen Gall
    • Computer Science
    • 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
    • 2020
    Enhancing Temporal Action Localization with Transfer Learning from Action Recognition
    Weakly Supervised Action Segmentation Using Mutual Consistency
    • 3
    • PDF
    Weakly Supervised Energy-Based Learning for Action Segmentation
    Weakly Supervised Energy-Based Learning for Action Segmentation
    • 9
    • Highly Influenced
    • PDF
    Weakly Supervised Energy-Based Learning for Action Segmentation
    Weakly Supervised Energy-Based Learning for Action Segmentation

    References

    SHOWING 1-10 OF 37 REFERENCES
    Connectionist Temporal Modeling for Weakly Supervised Action Labeling
    • 129
    • Highly Influential
    • PDF
    Weakly Supervised Action Learning with RNN Based Fine-to-Coarse Modeling
    • 83
    • PDF
    Temporal Segment Networks: Towards Good Practices for Deep Action Recognition
    • 1,537
    • PDF
    Two-Stream Convolutional Networks for Action Recognition in Videos
    • 4,088
    • PDF
    Temporal Action Detection with Structured Segment Networks
    • 229
    • PDF
    An end-to-end generative framework for video segmentation and recognition
    • 79
    • PDF
    UntrimmedNets for Weakly Supervised Action Recognition and Detection
    • 197
    • PDF
    You Lead, We Exceed: Labor-Free Video Concept Learning by Jointly Exploiting Web Videos and Images
    • 77
    • PDF
    Learning realistic human actions from movies
    • 3,375
    • PDF
    Temporal Localization of Fine-Grained Actions in Videos by Domain Transfer from Web Images
    • 96
    • PDF