Evaluation of Deep Convolutional Nets for Document Image Classification and Retrieval
It addresses document analysis for researchers and practitioners, showing CNNs' effectiveness but is incremental as it adapts existing methods to a new domain.
This paper tackled document image classification and retrieval by applying deep convolutional neural networks (CNNs), achieving a new state-of-the-art with features that outperform hand-crafted alternatives and are robust to compression.
This paper presents a new state-of-the-art for document image classification and retrieval, using features learned by deep convolutional neural networks (CNNs). In object and scene analysis, deep neural nets are capable of learning a hierarchical chain of abstraction from pixel inputs to concise and descriptive representations. The current work explores this capacity in the realm of document analysis, and confirms that this representation strategy is superior to a variety of popular hand-crafted alternatives. Experiments also show that (i) features extracted from CNNs are robust to compression, (ii) CNNs trained on non-document images transfer well to document analysis tasks, and (iii) enforcing region-specific feature-learning is unnecessary given sufficient training data. This work also makes available a new labelled subset of the IIT-CDIP collection, containing 400,000 document images across 16 categories, useful for training new CNNs for document analysis.