Model card
TrOCR-large 558M.
UnknownunknownUnknown paramsUnknown
Imported from Papers With Code
§ 01 · Benchmarks
Every benchmark TrOCR-large 558M has a recorded score for.
| # | Benchmark | Area · Task | Metric | Value | Rank | Date | Source |
|---|---|---|---|---|---|---|---|
| 01 | icdar-2013 | Computer Vision · Scene Text Detection | accuracy | 98.4% | #8 | 2021-09-21 | source ↗ |
| 02 | svt | Computer Vision · Scene Text Recognition | accuracy | 96.1% | #11 | 2021-09-21 | source ↗ |
| 03 | IAM | Computer Vision · Handwriting Recognition | cer | 2.9% | #17 | 2021-09-21 | source ↗ |
| 04 | svtp | Computer Vision · Scene Text Recognition | accuracy | 88.1% | #18 | 2021-09-21 | source ↗ |
| 05 | cute80 | Computer Vision · Scene Text Recognition | accuracy | 84.1% | #19 | 2021-09-21 | source ↗ |
| 06 | iiit5k | Computer Vision · Scene Text Recognition | accuracy | 94.1% | #19 | 2021-09-21 | source ↗ |
Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.
§ 02 · Strengths by area
Where TrOCR-large 558M actually performs.
§ 03 · Papers
2 papers with results for TrOCR-large 558M.
- 2021-09-21· Computer Vision· 5 results
TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models
- 2021-09-21· Computer Vision· 1 result
TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models
§ 04 · Related models
Other Unknown models scored on Codesota.
fglihai
Unknown params · 6 results · 1 SOTA
CLIP4STR-L
Unknown params · 1 result · 1 SOTA
USYD NLP_CS29-2
Unknown params · 6 results
Corner-based Region Proposals
Unknown params · 3 results
EAST + VGG16
Unknown params · 3 results
SSTD
Unknown params · 3 results
TextBoxes++_MS
Unknown params · 3 results
WordSup (VGG16-synth-coco)
Unknown params · 3 results
§ 05 · Sources & freshness
Where these numbers come from.
arxiv
5
results
papers-with-code
1
result
6 of 6 rows marked verified.