Home/Browse/coco-text

coco-text

Unknown

coco-text is a state-of-the-art machine learning benchmark indexed on Codesota. This page tracks published model results, top scores per metric, and the SOTA timeline for coco-text.

Benchmark Stats

Models12
Papers13
Metrics1

SOTA History

Not enough data to show trend.

f-measure

Higher is better

RankModelSourceScoreYearPaper
1TCM

CLIP-based detector with joint-dataset training. IJCAI 2025.

Editorial65.92026Source
2PANet (Joint)

Pixel Aggregation Network with joint-dataset fine-tuning.

Editorial64.52026Source
3LRANet

Low-Rank Approximation Network. AAAI 2024 Oral.

Editorial61.72026Source
4DPText-DETR

DETR-based with dynamic point queries. Joint training. AAAI 2023.

Editorial61.62026Source
5MAEDet

MAE-based self-supervised pretraining for text detection. IJCAI 2025.

Editorial60.62026Source
6DBNet

Differentiable Binarization with fine-tuning. AAAI 2020.

Editorial60.52026Source
7DBNet++

DB with Adaptive Scale Fusion. Joint training. TPAMI 2022.

Editorial59.52026Source
8SRFormer

Segmentation+Regression Transformer. AAAI 2024.

Editorial59.42026Source
9FCENet

Fourier Contour Embedding. CVPR 2021.

Editorial57.92026Source
10PSENet

Progressive Scale Expansion Network. CVPR 2019.

Editorial562026Source
11ABCNet v2

Adaptive Bezier-Curve Network v2. TPAMI 2021.

Editorial53.22026Source
12ABCNet v2

Adaptive Bezier-Curve Network v2. TPAMI 2021.

Editorial53.22026Source
13DRRG

Deep Relational Reasoning Graph. CVPR 2020.

Editorial31.92026Source

Submit a Result