Exploring the Challenges towards Lifelong Fact Learning

@inproceedings{Elhoseiny2018ExploringTC,
  title={Exploring the Challenges towards Lifelong Fact Learning},
  author={Mohamed Elhoseiny and F. Babiloni and Rahaf Aljundi and Marcus Rohrbach and Manohar Paluri and T. Tuytelaars},
  booktitle={ACCV},
  year={2018}
}
So far life-long learning (LLL) has been studied in relatively small-scale and relatively artificial setups. Here, we introduce a new large-scale alternative. What makes the proposed setup more natural and closer to human-like visual systems is threefold: First, we focus on concepts (or facts, as we call them) of varying complexity, ranging from single objects to more complex structures such as objects performing actions, and objects interacting with other objects. Second, as in real-world… Expand
10 Citations
Task-Free Continual Learning
  • 81
  • PDF
Normalization Matters in Zero-Shot Learning
  • 6
  • PDF
LAMAL: LAnguage Modeling Is All You Need for Lifelong Language Learning
  • 6
  • PDF
LAMAL: LANGUAGE MODELING IS ALL YOU NEED
  • 2019
LAMOL: LAnguage MOdeling for Lifelong Language Learning
  • 22
Does Continual Learning = Catastrophic Forgetting?
  • PDF
Continual Learning in Neural Networks
  • 7
  • PDF
Compositional Continual Language Learning
  • 4

References

SHOWING 1-10 OF 31 REFERENCES
Sherlock: Scalable Fact Learning in Images
  • 17
  • PDF
Learning Everything about Anything: Webly-Supervised Visual Concept Learning
  • 268
  • PDF
Large-Scale Visual Relationship Understanding
  • 88
  • PDF
Riemannian Walk for Incremental Learning: Understanding Forgetting and Intransigence
  • 230
  • PDF
An embarrassingly simple approach to zero-shot learning
  • 715
  • PDF
Gradient Episodic Memory for Continual Learning
  • 614
  • PDF
Encoder Based Lifelong Learning
  • 119
  • PDF
Expert Gate: Lifelong Learning with a Network of Experts
  • 161
  • PDF
An Empirical Study and Analysis of Generalized Zero-Shot Learning for Object Recognition in the Wild
  • 286
  • PDF
Never-Ending Learning
  • 583
  • PDF
...
1
2
3
4
...