# Back-Tracking from Clarity: Self-Learning to See Text from Afar

arXiv: https://arxiv.org/abs/2609.32477 · HTML: https://mlsanity.com/paper/2609.32477/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 27 of its 46 references that are in the map.
- First posted on arXiv 26 Sep 2026.

## Builds on (most cited first, 27 of 27)

- [Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection](https://mlsanity.com/paper/2303.05499/) (2023)
- [SAM 2: Segment Anything in Images and Videos](https://mlsanity.com/paper/2408.00714/) (2024)
- [Self-training with Noisy Student improves ImageNet classification](https://mlsanity.com/paper/1911.04252/) (2019)
- [ByteTrack: Multi-Object Tracking by Associating Every Detection Box](https://mlsanity.com/paper/2110.06864/) (2021)
- [Fully-Convolutional Siamese Networks for Object Tracking](https://mlsanity.com/paper/1606.09549/) (2016)
- [SiamRPN++: Evolution of Siamese Visual Tracking with Very Deep Networks](https://mlsanity.com/paper/1812.11703/) (2018)
- [LaSOT: A High-quality Benchmark for Large-scale Single Object Tracking](https://mlsanity.com/paper/1809.07845/) (2018)
- [Synthetic Data for Text Localisation in Natural Images](https://mlsanity.com/paper/1604.06646/) (2016)
- [Learning Spatio-Temporal Transformer for Visual Tracking](https://mlsanity.com/paper/2103.17154/) (2021)
- [MixFormer: End-to-End Tracking with Iterative Mixed Attention](https://mlsanity.com/paper/2203.11082/) (2022)
- [CoTracker3: Simpler and Better Point Tracking by Pseudo-Labelling Real Videos](https://mlsanity.com/paper/2410.11831/) (2024)
- [ICDAR2017 Competition on Reading Chinese Text in the Wild (RCTW-17)](https://mlsanity.com/paper/1708.09585/) (2017)
- [Perceptual Generative Adversarial Networks for Small Object Detection](https://mlsanity.com/paper/1706.05274/) (2017)
- [DeepSolo: Let Transformer Decoder with Explicit Points Solo for Text Spotting](https://mlsanity.com/paper/2211.10772/) (2022)
- [ARTrackV2: Prompting Autoregressive Tracker Where to Look and How to Describe](https://mlsanity.com/paper/2312.17133/) (2023)
- [DPText-DETR: Towards Better Scene Text Detection with Dynamic Points in Transformer](https://mlsanity.com/paper/2207.04491/) (2022)
- [Multi-Oriented Text Detection with Fully Convolutional Networks](https://mlsanity.com/paper/1604.04018/) (2016)
- [A Bilingual, OpenWorld Video Text Dataset and End-to-end Video Text Spotter with Transformer](https://mlsanity.com/paper/2112.04888/) (2021)
- [RoadText-1K: Text Detection & Recognition Dataset for Driving Videos](https://mlsanity.com/paper/2005.09496/) (2020)
- [End-to-End Video Text Spotting with Transformer](https://mlsanity.com/paper/2203.10539/) (2022)
- [Small Object Detection: A Comprehensive Survey on Challenges, Techniques and Real-World Applications](https://mlsanity.com/paper/2503.20516/) (2025)
- [GoMatching: A Simple Baseline for Video Text Spotting via Long and Short Term Matching](https://mlsanity.com/paper/2401.07080/) (2024)
- [ICDAR 2015 Text Reading in the Wild Competition](https://mlsanity.com/paper/1506.03184/) (2015)
- [Minimum Delay Object Detection From Video](https://mlsanity.com/paper/1908.11092/) (2019)
- [Knowledge Distillation for Human Action Anticipation](https://mlsanity.com/paper/1904.04868/) (2019)
- [DSText V2: A Comprehensive Video Text Spotting Dataset for Dense and Small Text](https://mlsanity.com/paper/2312.01938/) (2023)
- [Text Detection & Recognition in the Wild for Robot Localization](https://mlsanity.com/paper/2205.08565/) (2022)

## Abstract (arXiv)

We propose a self-supervised framework designed to enhance the capability of scene text detectors in identifying and recognizing text in scenarios where instances are shown at significant distances, typically small, blurred, and frequently missed by conventional models. Our approach leverages the high-fidelity performance of existing text spotting models on large, clear text as a foundational supervisor. By temporally back-tracking these high-confidence detections through video sequences, we automatically synthesize pseudo-labels for preceding frames where the distant text is still visually degraded or undersized. These pseudo-labels enable training a student model specialized for early text detection, without requiring any manual annotation. The success of this approach depends on accurate pseudo-label generation, for which we develop a dedicated scene text tracker capable of maintaining consistent text identities across challenging video sequences. In addition, we propose SceneText50, a diverse multilingual outdoor dataset to facilitate training and evaluation. Experiments show that our framework significantly improves early detection accuracy and robustness across varied scenes and languages. Code and data are at hrefhttps://github.com/trid2912/BackTrackingTexthttps://github.com/trid2912/BackTrackingText.