Learning task-agnostic representation via toddler-inspired learning

TL;DR

This paper introduces a toddler-inspired interactive learning agent that develops task-agnostic visual representations, enabling better generalization across multiple vision tasks compared to traditional autoencoder-based models.

Contribution

The paper presents a novel interactive learning framework inspired by toddlers, which learns versatile visual representations without task-specific supervision.

Findings

Achieved 100% accuracy in image classification

Attained 75.1% accuracy in object localization

Reached 1.62% relative error in distance estimation

Abstract

One of the inherent limitations of current AI systems, stemming from the passive learning mechanisms (e.g., supervised learning), is that they perform well on labeled datasets but cannot deduce knowledge on their own. To tackle this problem, we derive inspiration from a highly intentional learning system via action: the toddler. Inspired by the toddler's learning procedure, we design an interactive agent that can learn and store task-agnostic visual representation while exploring and interacting with objects in the virtual environment. Experimental results show that such obtained representation was expandable to various vision tasks such as image classification, object localization, and distance estimation tasks. In specific, the proposed model achieved 100%, 75.1% accuracy and 1.62% relative error, respectively, which is noticeably better than autoencoder-based model (99.7%, 66.1%, 1.95%), and also comparable with those of supervised models (100%, 87.3%, 0.71%).