Leveraging Semantic Embeddings for Safety-Critical Applications

Autor:	Brunner, Thomas, Diehl, Frederik, Le, Michael Truong, Knoll, Alois
Rok vydání:	2019
Předmět:	Statistics - Machine Learning Computer Science - Computer Vision and Pattern Recognition Computer Science - Machine Learning
Druh dokumentu:	Working Paper
Popis:	Semantic Embeddings are a popular way to represent knowledge in the field of zero-shot learning. We observe their interpretability and discuss their potential utility in a safety-critical context. Concretely, we propose to use them to add introspection and error detection capabilities to neural network classifiers. First, we show how to create embeddings from symbolic domain knowledge. We discuss how to use them for interpreting mispredictions and propose a simple error detection scheme. We then introduce the concept of semantic distance: a real-valued score that measures confidence in the semantic space. We evaluate this score on a traffic sign classifier and find that it achieves near state-of-the-art performance, while being significantly faster to compute than other confidence scores. Our approach requires no changes to the original network and is thus applicable to any task for which domain knowledge is available. Comment: Accepted at CVPR 2019 Workshop: Safe Artificial Intelligence for Automated Driving
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/1905.07733 Zobrazit plný text záznamu View this record from Arxiv