SynthTIGER: Synthetic Text Image GEneratoR Towards Better Text Recognition Models
This work addresses the lack of annotated real-world text images for training scene text recognition models, offering an incremental improvement in synthetic data generation.
The paper tackles the problem of generating synthetic text images to improve scene text recognition (STR) models by introducing SynthTIGER, which integrates effective synthesis techniques and addresses long-tail distributions in training data. It achieves better STR performance than existing synthetic datasets like MJSynth and SynthText, as demonstrated in experiments.
For successful scene text recognition (STR) models, synthetic text image generators have alleviated the lack of annotated text images from the real world. Specifically, they generate multiple text images with diverse backgrounds, font styles, and text shapes and enable STR models to learn visual patterns that might not be accessible from manually annotated data. In this paper, we introduce a new synthetic text image generator, SynthTIGER, by analyzing techniques used for text image synthesis and integrating effective ones under a single algorithm. Moreover, we propose two techniques that alleviate the long-tail problem in length and character distributions of training data. In our experiments, SynthTIGER achieves better STR performance than the combination of synthetic datasets, MJSynth (MJ) and SynthText (ST). Our ablation study demonstrates the benefits of using sub-components of SynthTIGER and the guideline on generating synthetic text images for STR models. Our implementation is publicly available at https://github.com/clovaai/synthtiger.