Corpus ID: 210700123

End-to-End Pixel-Based Deep Active Inference for Body Perception and Action

@article{Sancaktar2020EndtoEndPD,
  title={End-to-End Pixel-Based Deep Active Inference for Body Perception and Action},
  author={Cansu Sancaktar and Pablo Lanillos},
  journal={ArXiv},
  year={2020},
  volume={abs/2001.05847}
}
  • Cansu Sancaktar, Pablo Lanillos
  • Published 2020
  • Computer Science, Biology
  • ArXiv
  • We present a pixel-based deep active inference algorithm (PixelAI) inspired by human body perception and action. Our algorithm combines the free-energy principle from neuroscience, rooted in variational inference, with deep convolutional decoders to scale the algorithm to directly deal with raw visual input and provide online adaptive inference. Our approach is validated by studying body perception and action in a simulated and a real Nao robot. Results show that our approach allows the robot… CONTINUE READING

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 32 REFERENCES

    Deep Active Inference as Variational Policy Gradients

    VIEW 1 EXCERPT

    End-to-End Training of Deep Visuomotor Policies

    VIEW 1 EXCERPT

    Yielding Self-Perception in Robots Through Sensorimotor Contingencies

    VIEW 1 EXCERPT

    Scaling active inference

    VIEW 1 EXCERPT

    Visual Reinforcement Learning with Imagined Goals

    VIEW 1 EXCERPT

    Drifting perceptual patterns suggest prediction errors fusion rather than hypothesis selection: replicating the rubber-hand illusion on a robot