• Corpus ID: 37928675

Toward predictive machine learning for active vision

  title={Toward predictive machine learning for active vision},
  author={Emmanuel Dauc{\'e}},
  • E. Daucé
  • Published 28 October 2017
  • Computer Science
  • ArXiv
We develop a comprehensive description of the active inference framework, as proposed by Friston (2010), under a machine-learning compliant perspective. Stemming from a biological inspiration and the auto-encoding principles, the sketch of a cognitive architecture is proposed that should provide ways to implement estimation-oriented control policies. Computer simulations illustrate the effectiveness of the approach through a foveated inspection of the input data. The pros and cons of the… 

Figures from this paper

Hybrid Active Inference

It is suggested that, in addition to enhancing human cognitive functions with an intelligent and adaptive interface, integrated cognitive processing could accelerate emergent properties within artificial intelligence.

Not Only Look But Observe: Variational Observation Model of Scene-Level 3D Multi-Object Understanding for Probabilistic SLAM

This work proposes a method to approximate the Bayesian observation model of scene-level 3D multi-object understanding by exploiting variational auto-encoder (VAE), and estimates latent variables from the entire scene, which follow tractable distributions and concurrently imply 3D full shape and pose.

Enhancing geometric maps through environmental interactions

The deployment of rescue robots in real operations is becoming increasingly commonthanks to recent advances in AI technologies and high performance hardware. Rescue robots can now operate for exten



Infomax Control of Eye Movements

It is shown how an optimal eye movement controller can be learned from subjective experiences of information gathering, and in simulation properties of the optimal controller this controller outperforms other eye movement strategies proposed in the literature.

Computational modelling of visual attention

Five important trends have emerged from recent work on computational models of focal visual attention that emphasize the bottom-up, image-based control of attentional deployment, providing a framework for a computational and neurobiological understanding of visual attention.

Predictive coding in the visual cortex: a functional interpretation of some extra-classical receptive-field effects.

Results suggest that rather than being exclusively feedforward phenomena, nonclassical surround effects in the visual cortex may also result from cortico-cortical feedback as a consequence of the visual system using an efficient hierarchical strategy for encoding natural images.

Active perception

This thesis explores a robust mathematical foundation for understanding active perception, and develops many techniques for practical analysis for building machines that operate in real time in the real world, using actions to aid perception.

Bayesian Compression for Deep Learning

This work argues that the most principled and effective way to attack the problem of compression and computational efficiency in deep learning is by adopting a Bayesian point of view, where through sparsity inducing priors the authors prune large parts of the network.

Active Sensing as Bayes-Optimal Sequential Decision Making

This work presents a Bayes-optimal inference and control framework for active sensing, C-DAC (Context-Dependent Active Controller), and presents both parametric and non-parametric approximations, which retain context-sensitivity while significantly reducing computational complexity.

Perceptions as Hypotheses: Saccades as Experiments

This work explores the idea that saccadic eye movements are optimal experiments, in which data are gathered to test hypotheses or beliefs about how those data are caused, and provides a plausible model of visual search that can be motivated from the basic principles of self-organized behavior.

Optimal eye movement strategies in visual search

This work derives the ideal bayesian observer for search tasks in which a target is embedded at an unknown location within a random background that has the spectral characteristics of natural scenes and finds that humans achieve nearly optimal search performance, even though humans integrate information poorly across fixations.

Active vision

It is proved that an active observer can solve basic vision problems in a much more efficient way than a passive one, and a general methodology is introduced, a general framework in which low-level vision problems should be addressed.

Metamers of the ventral stream

A population model for mid-ventral processing is developed, in which nonlinear combinations of V1 responses are averaged in receptive fields that grow with eccentricity, providing a quantitative framework for assessing the capabilities and limitations of everyday vision.