From phonemes to images: levels of representation in a recurrent neural model of visually-grounded language learning

Abstract

We present a model of visually-grounded language learning based on stacked gated recurrent neural networks which learns to predict visual features given an image description in the form of a sequence of phonemes. The learning task resembles that faced by human language learners who need to discover both structure and meaning from noisy and ambiguous data across modalities. We show that our model indeed learns to predict features of the visual context given phonetically transcribed image descriptions, and show that it represents linguistic information in a hierarchy of levels: lower layers in the stack are comparatively more sensitive to form, whereas higher layers are more sensitive to meaning.

Accepted at COLING 2016

References in corpus (3)

• Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling
• Theano: new features and speed improvements
• Skip-Thought Vectors

Cited by in corpus (6)

• The emergence of number and syntax units in LSTM language models
• Visually grounded models of spoken language: A survey of datasets, architectures and evaluation techniques
• Analyzing Hidden Representations in End-to-End Automatic Speech Recognition Systems
• Talk, Don't Write: A Study of Direct Speech-Based Image Retrieval
• Analysis Methods in Neural Language Processing: A Survey
• Towards localisation of keywords in speech using weak supervision