Total-Text: A Comprehensive Dataset for Scene Text Detection and Recognition

By Chee-Kheng Chng, Chee Seng Chan. IEEE International Conference on Document Analysis and Recognition, 2017

542 citations recorded by the source corpus.

Text in curve orientation, despite being one of the common text orientations in real world environment, has close to zero existence in well received scene text datasets such as ICDAR'13 and MSRA-TD500. The main motivation of Total-Text is to fill this gap and facilitate a new research direction for the scene text community. On top of conventional horizontal and multi-oriented text, it features curved-oriented text. Total-Text is highly diversified in orientations, more than half of its images have a combination of more than two orientations. Recently, a new breed of solutions that casted text detection as a segmentation problem has demonstrated their effectiveness against multi-oriented text. In order to evaluate its robustness against curved text, we fine-tuned DeconvNet and benchmark it on Total-Text. Total-Text with its annotation is available at https://github.com/cs-chan/Total-Text-Dataset.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/729f20fa5928514768f4e0d848e74e5eecc33174

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

Character Region Awareness for Text Detection (2019, methodology): “Furthermore, experiments on MSRA-TD500, CTW-1500, and TotalText datasets [36, 38, 3] show the high ﬂexibility of the proposed method on complicated cases, such as long, curved, and/or arbitrarily shaped texts.”
TextSnake: A Flexible Representation for Detecting Text of Arbitrary Shapes (2018, methodology): “TotalText [12] is a newly-released benchmark for text detection.”
Look More Than Once: An Accurate Detector for Text of Arbitrary Shapes (2019, methodology): “The Total-Text [2] is another curved text benchmarks, which consist of 1255 training images and 300 testing images.”
ABCNet: Real-Time Scene Text Spotting With Adaptive Bezier-Curve Network (2020, methodology): “For Total-Text and CTW1500, m is 5 and 7, respectively. t is calculated by using the ratio of the cumulative length to the perimeter of the polyline.”
ImageBind-LLM: Multi-modality Instruction Tuning (2023, methodology): “We evaluate ImageBind-LLM on 12 representative OCR datasets, including IIIT5K [61], ICDAR 2013(IC13) [62], ICDAR 2015 (IC15) [63], Total-Text [64], CUTE80 [65], Street View Text (SVT) [66], SVTP-Perspective (SVTP) [67], COCO-Text [68], WordArt [69], SCUT-CTW1500 (CTW) [70], Heavily Occluded Scene Te”
YOLO-based Object Detection Models: A Review and its Applications (2024, unknown): “Other datasets that are used in previous research work include Remote sensing target detection [42] [43] [44], text detection [45] [46] [47], and Pedestrian detection [48] [49] [50].”
Transformers in Remote Sensing: A Survey (2022, methodology): “RBox [42] Detection SynthText [89], ICDAR 2015 (IC15) [90], MLT-2017 (MLT17) [91], MSRA-TD500 [92], MTWI [93], Total-Text [94], CTW1500 [95] mAP Proposes a framework employing transformers to model the relationship of sampled features for better grouping and box prediction without requiring post-pro”
Real-time Scene Text Detection with Differentiable Binarization (2019, background): “Total-Text dataset Total-Text (Ch’ng and Chan 2017) is a dataset that includes the text of various shapes, including horizontal, multi-oriented, and curved.”
Scene Text Detection and Recognition: The Deep Learning Era (2018, methodology): “For example, there are datasets featuring long text (Tu et al. 2012), blurred text (Karatzas et al. 2015), and curved text (Ch’ng and Chan 2017) respectively.”
Shape Robust Text Detection With Progressive Scale Expansion Network (2018, methodology): “To show the effectiveness of our proposed PSENet, we conduct extensive experiments on four competitive benchmark datasets including ICDAR 2015 [17], ICDAR 2017 MLT [1] ,CTW1500 [24] and Total-Text [2].”
Efficient and Accurate Arbitrary-Shaped Text Detection With Pixel Aggregation Network (2019, methodology): “Extensive experiments on Total-Text and CTW1500 demonstrate the superior advantages in speed and accuracy when compared to previous state-of-the-art text detectors.”
Total-Text: toward orientation robustness in scene text detection (2019, methodology): “Firstly, we address the inconsistency of the polygon annotation employed in our preliminary version [12] in Sect.”
OCRBench: on the hidden mystery of OCR in large multimodal models (2023, unknown): “…SVT [38], IC13 [39]; (2)Irregular Text Recognition: IC15 [40], SVTP [41], CT80 [42], COCOText(COCO) [43], SCUT-CTW1500 (CTW) [44], Total-Text (TT) [45]; (3) Occlusion Scene Text [46], encompassing weakly occluded scene text (WOST) and heavily occluded scene text (HOST); (4 LMMs primarily rely on…”
Structured Multimodal Attentions for TextVQA (2020, methodology): “The SBD model is pretrained on a 60 k dataset, which consists of 30 , 000 images from LSVT [56] training set, 10 , 000 images from MLT 2019 [57] training set, 5 , 603 images from ArT [58] (containing all the images of SCUT-CTW1500 [59] and Total-text [60], [61]), and 14 , 859 images selected from a ”
Sliding Line Point Regression for Shape Robust Scene Text Detection (2018, background): “Therefore, recent research [13] [14] have begun to explore curve text line detection.”
LVLM-EHub: A Comprehensive Evaluation Benchmark for Large Vision-Language Models (2023, unknown): “We hope that our conclusions can pave the way for developing innovative strategies aimed at enhancing multimodal techniques.”