Tracking Based Semi-Automatic Annotation for Scene Text Videos

被引:2
|
作者
Zhu, Jiajun [1 ]
Jiang, Xiufeng [1 ]
Jia, Zhiwei [1 ]
Xu, Shugong [1 ]
Cao, Shan [1 ]
机构
[1] Shanghai Univ, Shanghai Inst Adv Commun & Data Sci, Shanghai 200444, Peoples R China
关键词
Annotations; Labeling; Detectors; Manuals; Correlation; Visualization; Task analysis; Scene text video; tracking; semi-automatic labeling;
D O I
10.1109/ACCESS.2021.3066601
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Recently, video scene text detection has received increasing attention due to its comprehensive applications. However, the lack of annotated scene text video datasets has become one of the most important problems, which hinders the development of video scene text detection. The existing scene text video datasets are not large-scale due to the expensive cost caused by manual labeling. In addition, the text instances in these datasets are too clear to be a challenge. To address the above issues, we propose a tracking based semi-automatic labeling strategy for scene text videos in this paper. We get semi-automatic scene text annotation by labeling manually for the first frame and tracking automatically for the subsequent frames, which avoid the huge cost of manual labeling. Moreover, a paired low-quality scene text video dataset named Text-RBL is proposed, consisting of raw videos, blurry videos, and low-resolution videos, labeled by the proposed convenient semi-automatic labeling strategy. Through an averaging operation and bicubic down-sampling operation over the raw videos, we can efficiently obtain blurry videos and low-resolution videos paired with raw videos separately. To verify the effectiveness of Text-RBL, we propose a baseline model combined with the text detector and tracker for video scene text detection. Moreover, a failure detection scheme is designed to alleviate the baseline model drift issue caused by complex scenes. Extensive experiments demonstrate that Text-RBL with paired low-quality videos labeled by the semi-automatic method can significantly improve the performance of the text detector in low-quality scenes.
引用
收藏
页码:46325 / 46338
页数:14
相关论文
共 50 条
  • [41] Multi-Strategy Tracking Based Text Detection in Scene Videos
    Zuo, Ze-Yu
    Tian, Shu
    Pei, Wei-yi
    Yin, Xu-Cheng
    2015 13TH IAPR INTERNATIONAL CONFERENCE ON DOCUMENT ANALYSIS AND RECOGNITION (ICDAR), 2015, : 66 - 70
  • [42] Assis: Online Semi-Automatic Dialog Annotation Tool
    Schutz Foerste, Henrique Theodor
    Malta Purim, Andreis Gustavo
    de Souza, Rafael Roque
    dos Reis, Julio Cesar
    PROCEEDINGS OF THE 19TH BRAZILIAN SYMPOSIUM ON INFORMATION SYSTEMS, 2023, : 37 - 44
  • [43] Semi-automatic Annotation of OCT Images for CNN Training
    Schleier, Sebastian
    Stolz, Noah
    Langner, Holger
    Hasan, Rama
    Roschke, Christian
    Ritter, Marc
    HUMAN-COMPUTER INTERACTION. DESIGN AND USER EXPERIENCE, HCI 2020, PT I, 2020, 12181 : 672 - 685
  • [44] A semi-automatic annotation tool for unobtrusive gesture analysis
    Stijn De Beugher
    Geert Brône
    Toon Goedemé
    Language Resources and Evaluation, 2018, 52 : 433 - 460
  • [45] A semiotic semi-automatic annotation for movie audiovisual document
    Fourati, Manel
    Chaari, Abir
    Jedidi, Anis
    Gargouri, Faiez
    2015 15TH INTERNATIONAL CONFERENCE ON INTELLIGENT SYSTEMS DESIGN AND APPLICATIONS (ISDA), 2015, : 533 - 539
  • [46] Visual OntoBridge: Semi-automatic Semantic Annotation Software
    Grcar, Miha
    Mladenic, Dunja
    MACHINE LEARNING AND KNOWLEDGE DISCOVERY IN DATABASES, PT II, 2009, 5782 : 726 - 729
  • [47] A MATHEMATICAL APPROACH TOWARDS SEMI-AUTOMATIC IMAGE ANNOTATION
    Seneviratne, L.
    Izquierdo, E.
    19TH EUROPEAN SIGNAL PROCESSING CONFERENCE (EUSIPCO-2011), 2011, : 559 - 563
  • [48] A Semi-Automatic Annotation Approach for Human Activity Recognition
    Bota, Patricia
    Silva, Joana
    Folgado, Duarte
    Gamboa, Hugo
    SENSORS, 2019, 19 (03):
  • [49] Supporting semi-automatic semantic annotation of multimedia resources
    Pan, Jeff Z.
    Mellish, Chris
    ARTIFICIAL INTELLIGENCE APPLICATIONS AND INNOVATIONS, 2006, 204 : 609 - +
  • [50] Semi-automatic landmark point annotation for geometric morphometrics
    Paul A Bromiley
    Anja C Schunke
    Hossein Ragheb
    Neil A Thacker
    Diethard Tautz
    Frontiers in Zoology, 11