Papers › DTrOCR: Decoder-only Transformer for Optical Character Recognition
DTrOCR: Decoder-only Transformer for Optical Character Recognition
Masato Fujitake
Typical text recognition methods rely on an encoder-decoder structure, in which the encoder extracts features from an image, and the decoder produces recognized text from these features. In this study, we propose a simpler and more effective method for text recognition, known as the Decoder-only Transformer for Optical Character Recognition (DTrOCR). This method uses a decoder-only Transformer to take advantage of a generative language model that is pre-trained on a large corpus. We examined whether a generative language model that has been successful in natural language processing can also be effective for text recognition in computer vision. Our experiments demonstrated that DTrOCR outperforms current state-of-the-art methods by a large margin in the recognition of printed, handwritten, and scene text in both English and Chinese.
In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.
Code
Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.
Code Syntology ran Syntology
Not run by Syntology. Nothing on this page verifies that the listed code works.
Tasks
Results from the paper archive 2025-07-28
| Task | Dataset | Model | Metric | Value | Rank at snapshot | Leaderboard | Report |
|---|---|---|---|---|---|---|---|
| Handwritten Text Recognition | IAM | DTrOCR 105M | CER | 2.38 | #1 of 17 | Archive leaderboard | report |
| Optical Character Recognition (OCR) | Benchmarking Chinese Text Recognition: Datasets, Baselines, and an Empirical Study | DTrOCR | Accuracy (%) | 89.6 | #1 of 7 | Archive leaderboard | report |
| Optical Character Recognition (OCR) | Benchmarking Chinese Text Recognition: Datasets, Baselines, and an Empirical Study | DTrOCR 105M | Accuracy (%) | 89.6 | #2 of 7 | Archive leaderboard | report |
| Scene Text Recognition | CUTE80 | DTrOCR 105M | Accuracy | 99.1 | #6 of 18 | Archive leaderboard | report |
| Scene Text Recognition | ICDAR2013 | DTrOCR 105M | Accuracy | 99.4 | #2 of 38 | Archive leaderboard | report |
| Scene Text Recognition | ICDAR2015 | DTrOCR 105M | Accuracy | 93.5 | #1 of 27 | Archive leaderboard | report |
| Scene Text Recognition | IIIT5k | DTrOCR 105M | Accuracy | 99.6 | #2 of 17 | Archive leaderboard | report |
| Scene Text Recognition | SVT | DTrOCR 105M | Accuracy | 98.9 | #2 of 37 | Archive leaderboard | report |
| Scene Text Recognition | SVTP | DTrOCR 105M | Accuracy | 98.6 | #1 of 17 | Archive leaderboard | report |
Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.
Methods
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections