Synthetic Data for Text Localisation in Natural Images
arXiv:1604.06646
Abstract
In this paper we introduce a new method for text detection in natural images. The method comprises two contributions: First, a fast and scalable engine to generate synthetic images of text in clutter. This engine overlays synthetic text to existing background images in a natural way, accounting for the local 3D scene geometry. Second, we use the synthetic images to train a Fully-Convolutional Regression Network (FCRN) which efficiently performs text detection and bounding-box regression at all locations and multiple scales in an image. We discuss the relation of FCRN to the recently-introduced YOLO detector, as well as other end-to-end object detection systems based on deep learning. The resulting detection network significantly out performs current methods for text detection in natural images, achieving an F-measure of 84.2% on the standard ICDAR 2013 benchmark. Furthermore, it can process 15 images per second on a GPU.
References in corpus (3)
Cited by in corpus (69)
- TextBoxes++: A Single-Shot Oriented Scene Text Detector
- TextBoxes: A Fast Text Detector with a Single Deep Neural Network
- Modeling Surface Appearance from a Single Photograph using Self-augmented Convolutional Neural Networks
- Universal representations:The missing link between faces, text, planktons, and cat breeds
- EAST: An Efficient and Accurate Scene Text Detector
- Weak Supervision for Generating Pixel-Level Annotations in Scene Text Segmentation
- Single Shot Text Detector with Regional Attention
- Towards Accurate Scene Text Recognition with Semantic Reasoning Networks
- Deep Direct Regression for Multi-Oriented Scene Text Detection
- PixelLink: Detecting Scene Text via Instance Segmentation
- STN-OCR: A single Neural Network for Text Detection and Text Recognition
- Training Deep Face Recognition Systems with Synthetic Data
- WordSup: Exploiting Word Annotations for Character based Text Detection
- Multi-Oriented Scene Text Detection via Corner Localization and Region Segmentation
- Adversarial Generation of Training Examples: Applications to Moving Vehicle License Plate Recognition
- Scene Text Recognition from Two-Dimensional Perspective
- Scene Text Synthesis for Efficient and Effective Deep Network Training
- Edit Probability for Scene Text Recognition
- Understanding Infographics through Textual and Visual Tag Prediction
- Learning Markov Clustering Networks for Scene Text Detection
- Pixel-Anchor: A Fast Oriented Scene Text Detector with Combined Networks
- Sliding Line Point Regression for Shape Robust Scene Text Detection
- Fused Text Segmentation Networks for Multi-oriented Scene Text Detection
- Scene Text Detection with Supervised Pyramid Context Network
- A pooling based scene text proposal technique for scene text reading in the wild
- Rotation-Sensitive Regression for Oriented Scene Text Detection
- Feature Enhancement Network: A Refined Scene Text Detector
- Fast and Robust Multiple ColorChecker Detection using Deep Convolutional Neural Networks
- WeText: Scene Text Detection under Weak Supervision
- ResearchDoom and CocoDoom: Learning Computer Vision with Games
- Detecting Oriented Text in Natural Images by Linking Segments
- Modeling Visual Context is Key to Augmenting Object Detection Datasets
- A Multi-Object Rectified Attention Network for Scene Text Recognition
- Verisimilar Image Synthesis for Accurate Detection and Recognition of Texts in Scenes
- WordFence: Text Detection in Natural Images with Border Awareness
- Recurrent Calibration Network for Irregular Text Recognition
- ICDAR 2019 Robust Reading Challenge on Reading Chinese Text on Signboard
- Bidirectional Scene Text Recognition with a Single Decoder
- FOTS: Fast Oriented Text Spotting with a Unified Network
- Generating Text Sequence Images for Recognition
- Learning from scarce information: using synthetic data to classify Roman fine ware pottery
- Learning to detect and localize many objects from few examples
- ArbiText: Arbitrary-Oriented Text Detection in Unconstrained Scene
- Teaching Robots Novel Objects by Pointing at Them
- STRIDE : Scene Text Recognition In-Device
- Primitive Representation Learning for Scene Text Recognition
- All You Need Is Boundary: Toward Arbitrary-Shaped Text Spotting
- Dense Residual Network: Enhancing Global Dense Feature Flow for Character Recognition
- Text Perceptron: Towards End-to-End Arbitrary-Shaped Text Spotting
- Recovering Homography from Camera Captured Documents using Convolutional Neural Networks
- Learning to Read by Spelling: Towards Unsupervised Text Recognition
- Synthetically Trained Icon Proposals for Parsing and Summarizing Infographics
- Membership Inference Attacks Against Object Detection Models
- AON: Towards Arbitrarily-Oriented Text Recognition
- TextNet: Irregular Text Reading from Images with an End-to-End Trainable Network
- Cascaded Segmentation-Detection Networks for Word-Level Text Spotting
- Integrating Scene Text and Visual Appearance for Fine-Grained Image Classification
- Inductive Visual Localisation: Factorised Training for Superior Generalisation
- Simultaneous Recognition of Horizontal and Vertical Text in Natural Images
- Correlation Propagation Networks for Scene Text Detection
- Automated X-ray Image Analysis for Cargo Security: Critical Review and Future Promise
- STELA: A Real-Time Scene Text Detector with Learned Anchor
- EnsNet: Ensconce Text in the Wild
- Semantic Human Matting
- How many labeled license plates are needed?
- Forward Amortized Inference for Likelihood-Free Variational Marginalization
- Accurate Scene Text Detection through Border Semantics Awareness and Bootstrapping
- Exploring Font-independent Features for Scene Text Recognition
- SmartTennisTV: Automatic indexing of tennis videos