19 August 2025
Transfer Learning (TL) offers the potential to accelerate learning by transferring knowledge across tasks. However, it faces critical challenges such as negative transfer, domain adaptation and inefficiency in selecting solid source policies. These issues often represent critical problems in evolving domains, i.e. game development, where scenarios transform and agents must adapt. The continuous release of new agents is costly and inefficient.
In this work, EMERGE partners from the University of Pisa challenge the key issues in TL to improve knowledge transfer and agents' performance across tasks and reduce computational costs. The proposed methodology, called Framework for Adaptive Similarity-based Transfer (FAST), leverages visual frames and textual descriptions to create a latent representation of tasks dynamics, that is exploited to estimate similarity between environments. Experimental results, over multiple racing tracks, demonstrate that FAST achieves competitive final performance compared to learning-from-scratch methods while requiring significantly less training steps.
Read the paper in the link below.
More Information
