How much "human-like" visual experience do current self-supervised learning algorithms need to achieve human-level object recognition?

09/23/2021

∙

This paper addresses a fundamental question: how good are our current self-supervised visual representation learning algorithms relative to humans? More concretely, how much "human-like", natural visual experience would these algorithms need in order to reach human-level performance in a complex, realistic visual object recognition task such as ImageNet? Using a scaling experiment, here we estimate that the answer is on the order of a million years of natural visual experience, in other words several orders of magnitude longer than a human lifetime. However, this estimate is quite sensitive to some underlying assumptions, underscoring the need to run carefully controlled human experiments. We discuss the main caveats surrounding our estimate and the implications of this rather surprising result.

READ FULL TEXT

How much "human-like" visual experience do current self-supervised learning algorithms need to achieve human-level object recognition?

Scaling may be all you need for achieving human-level object recognition capacity with human-like visual experience

On the surprising similarities between supervised and self-supervised models

Reason from Context with Self-supervised Learning

The developmental trajectory of object recognition robustness: children are like small adults but unlike big deep neural networks

A degree of image identification at sub-human scales could be possible with more advanced clusters

Partial success in closing the gap between human and machine vision

GIFT: Generalizable Interaction-aware Functional Tool Affordances without Labels

How much "human-like" visual experience do current self-supervised learning algorithms need to achieve human-level object recognition?

Related Research

Scaling may be all you need for achieving human-level object recognition capacity with human-like visual experience

On the surprising similarities between supervised and self-supervised models

Reason from Context with Self-supervised Learning

The developmental trajectory of object recognition robustness: children are like small adults but unlike big deep neural networks

A degree of image identification at sub-human scales could be possible with more advanced clusters

Partial success in closing the gap between human and machine vision

GIFT: Generalizable Interaction-aware Functional Tool Affordances without Labels