Shared visual representations of drawing for communication: how do different biases affect human interpretability and intent?

Mihai, Andreea, Daniela and Hare, Jonathon (2021) Shared visual representations of drawing for communication: how do different biases affect human interpretability and intent? In Shared Visual Representations in Human and Machine Intelligence: 2021 NeurIPS Workshop. 10 pp .

Record type: Conference or Workshop Item (Paper)

Abstract

We present an investigation into how representational losses can affect the drawings produced by artificial agents playing a communication game. Building upon recent advances, we show that a combination of powerful pretrained encoder networks, with appropriate inductive biases, can lead to agents that draw recognisable sketches, whilst still communicating well. Further, we start to develop an approach to help automatically analyse the semantic content being conveyed by a sketch and demonstrate that current approaches to inducing perceptual biases lead to a notion of objectness being a key feature despite the agent training being self-supervised.

Text

SVRHM_2021___Sketching_with_CLIP (1)

Available under License University of Southampton Accepted Manuscript Licence.

Download (6MB)