FlowText: Synthesizing Realistic Scene Text Video with Optical Flow Estimation
This addresses the data scarcity issue for researchers and practitioners in video text analysis, though it is an incremental improvement over existing image-level synthesis methods.
The paper tackles the problem of expensive manual labeling for video text spotting by introducing FlowText, a low-cost synthetic data generation method that uses optical flow to create realistic scene text videos, achieving remarkable results on datasets like ICDAR2015video and ICDAR2013video.
Current video text spotting methods can achieve preferable performance, powered with sufficient labeled training data. However, labeling data manually is time-consuming and labor-intensive. To overcome this, using low-cost synthetic data is a promising alternative. This paper introduces a novel video text synthesis technique called FlowText, which utilizes optical flow estimation to synthesize a large amount of text video data at a low cost for training robust video text spotters. Unlike existing methods that focus on image-level synthesis, FlowText concentrates on synthesizing temporal information of text instances across consecutive frames using optical flow. This temporal information is crucial for accurately tracking and spotting text in video sequences, including text movement, distortion, appearance, disappearance, shelter, and blur. Experiments show that combining general detectors like TransDETR with the proposed FlowText produces remarkable results on various datasets, such as ICDAR2015video and ICDAR2013video. Code is available at https://github.com/callsys/FlowText.