Abstract:End-to-end text-spotting, which aims to integrate detection and recognition in a unified framework, has attracted increasing attention due to its simplicity of the two complimentary tasks. It remains an open problem especially when processing arbitrarily-shaped text instances. Previous methods can be roughly categorized into two groups: character-based and segmentation-based, which often require character-level annotations and/or complex post-processing due to the unstructured output. Here, we tackle end-to-end text spotting by presenting Adaptive Bezier Curve Network v2 (ABCNet v2). Our main contributions are four-fold: 1) For the first time, we adaptively fit arbitrarily-shaped text by a parameterized Bezier curve, which, compared with segmentation-based methods, can not only provide structured output but also controllable representation. 2) We design a novel BezierAlign layer for extracting accurate convolution features of a text instance of arbitrary shapes, significantly improving the precision of recognition over previous methods. 3) Different from previous methods, which often suffer from complex post-processing and sensitive hyper-parameters, our ABCNet v2 maintains a simple pipeline with the only post-processing non-maximum suppression (NMS). 4) As the performance of text recognition closely depends on feature alignment, ABCNet v2 further adopts a simple yet effective coordinate convolution to encode the position of the convolutional filters, which leads to a considerable improvement with negligible computation overhead. Comprehensive experiments conducted on various bilingual (English and Chinese) benchmark datasets demonstrate that ABCNet v2 can achieve state-of-the-art performance while maintaining very high efficiency. More importantly, as there is little work on quantization of text spotting models, we quantize our models to improve the inference time of the proposed ABCNet v2. This can be valuable for real-time applications. Code and model are availa-le at: https://git.io/AdelaiDet.

ABCNN: Attention-Based Convolutional Neural Network for Modeling Sentence Pairs.

Dual Edge-embedding Graph Convolutional Network for Unified Aspect-based Sentiment Analysis

Modelling Sentence Pairs with Tree-structured Attentive Encoder

Abircnn With Neural Tensor Network For Answer Selection

A Study on Automatic Recognition of Chinese Sentence Pairs Relations Based on CNN.

Natural Language Inference by Tree-Based Convolution and Heuristic Matching

A Convolutional Neural Network for Modelling Sentences

Learning Context-Sensitive Convolutional Filters for Text Processing

Attentive Convolution: Equipping CNNs with RNN-style Attention Mechanisms

A Sentence Interaction Network for Modeling Dependence Between Sentences.

A Multi-Size Neural Network with Attention Mechanism for Answer Selection

Recognizing Entailment and Contradiction by Tree-based Convolution.

Inter-Weighted Alignment Network for Sentence Pair Modeling.

Ercnn: Enhanced Recurrent Convolutional Neural Networks For Learning Sentence Similarity

Attentive Interactive Neural Networks for Answer Selection in Community Question Answering.

Modelling Interaction of Sentence Pair with Coupled-LSTMs.

Sentence-Embedding and Similarity via Hybrid Bidirectional-LSTM and CNN Utilizing Weighted-Pooling Attention

ABCNet v2: Adaptive Bezier-Curve Network for Real-time End-to-end Text Spotting

Convolutional Interaction Network for Natural Language Inference

Tree-based Convolution for Sentence Modeling.