Avoiding hashing and encouraging visual semantics in referential emergent language games
This work addresses the challenge of inducing meaningful visual semantics in multi-agent communication systems, which is incremental as it builds on existing game settings.
The paper tackled the problem of learning visual semantics in emergent communication games by investigating how feature extractor weights and task design affect the learned representations, demonstrating that self-supervised games can capture conceptual image properties.
There has been an increasing interest in the area of emergent communication between agents which learn to play referential signalling games with realistic images. In this work, we consider the signalling game setting of Havrylov and Titov and investigate the effect of the feature extractor's weights and of the task being solved on the visual semantics learned or captured by the models. We impose various augmentation to the input images and additional tasks in the game with the aim to induce visual representations which capture conceptual properties of images. Through our set of experiments, we demonstrate that communication systems which capture visual semantics can be learned in a completely self-supervised manner by playing the right types of game.