• Publications
  • Influence
Wronging a Right: Generating Better Errors to Improve Grammatical Error Detection
TLDR
We investigate cheaply constructing synthetic samples, given a small corpus of human-annotated data, using an off-the-rack attentive sequence-to-sequence model and a straight-forward post-processing procedure. Expand
  • 31
  • 7
  • PDF
Learning From Demonstration in the Wild
TLDR
We propose video to behaviour (ViBe), a new approach to learn models of behaviour from unlabelled raw video data of a traffic scene collected from a single, monocular, uncalibrated camera with ordinary resolution. Expand
  • 76
  • 3
  • PDF
Imagine That! Leveraging Emergent Affordances for 3D Tool Synthesis
TLDR
In this paper we explore the richness of information captured by the latent space of a vision-based generative model to learn and exploit task-relevant object affordances given visual observations from a reaching task, involving a scenario and a stick-like tool. Expand
  • 2
  • PDF
Imagine That! Leveraging Emergent Affordances for Tool Synthesis in Reaching Tasks
TLDR
In this paper we investigate an artificial agent's ability to perform task-focused tool synthesis via imagination. Expand
  • 2