Program Generation from Diverse Video Demonstrations

TL;DR

This paper introduces a novel multi-sequence-to-sequence model that learns to extract general rules from diverse video demonstrations, enabling program synthesis with improved accuracy and robustness to noise.

Contribution

The work presents a new approach that combines summarisation and translation in a multi-sequence-to-sequence framework for program generation from videos, outperforming prior methods.

Findings

Achieved 11.75% higher program accuracy on Vizdoom

Handled noisy specifications without additional filtering

Demonstrated state-of-the-art results in program synthesis

Abstract

The ability to use inductive reasoning to extract general rules from multiple observations is a vital indicator of intelligence. As humans, we use this ability to not only interpret the world around us, but also to predict the outcomes of the various interactions we experience. Generalising over multiple observations is a task that has historically presented difficulties for machines to grasp, especially when requiring computer vision. In this paper, we propose a model that can extract general rules from video demonstrations by simultaneously performing summarisation and translation. Our approach differs from prior works by framing the problem as a multi-sequence-to-sequence task, wherein summarisation is learnt by the model. This allows our model to utilise edge cases that would otherwise be suppressed or discarded by traditional summarisation techniques. Additionally, we show that our approach can handle noisy specifications without the need for additional filtering methods. We evaluate our model by synthesising programs from video demonstrations in the Vizdoom environment achieving state-of-the-art results with a relative increase of 11.75% program accuracy on prior works