A Deep-learning Model-based and Data-driven Hybrid Architecture for Image Annotation

摘要

Does adding more training data always help improve the effectiveness of a machine-learning or pattern-recognition task? Recent evidences in machine translation and speech recognition suggest that when abundant training data is available, the data-driven approach outperforms the traditional model-based approach. In this work, we compare representative data-driven and model-based schemes on an application of image annotation. We enumerate pros and cons of these two approaches, and propose a hybrid approach, which can harness the strengths of the two.