Scholarlink

This report presents the outcomes of the Summer Challenge on Writer Verification, hosted as part of the Eighth National Conference on Computer Vision, Pattern Recognition, Image Processing, and Graphics (NCVPRIPG) held at IIT Jodhpur on July 21–23, 2023. This challenge introduces a novel dataset comprising images of handwritten text contributed by 1,352 unique writers. Predominantly, these images feature handwritten Hindi text, but they also encompass a variety of elements such as numbers, mathematical symbols, and English text. Participants were tasked with developing a model capable of automatically determining whether the text in a given pair of images is authored by the same writer or different writers. The primary objective of this challenge was to advance research in the realm of handwritten text recognition. Throughout the competition, we registered 108 teams, with 18 teams submitting results for the validation dataset. Out of these, 13 teams provided submissions for the semi-final dataset. The top six teams from the semi-finals were subsequently invited to compete in the finals. Additional details about the challenge can be found at https://vl2g.github.io/challenges/wv2023/ .

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

NCVPRIPG 2023 Summer Challenge on Writer Verification

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Gyan Prabhat,
  • Abu Shahid,
  • Prajwal Gatti,
  • Manan Shah,
  • Girin Chutia,
  • Arvind Kumar Sharma,
  • Anand Mishra

摘要

This report presents the outcomes of the Summer Challenge on Writer Verification, hosted as part of the Eighth National Conference on Computer Vision, Pattern Recognition, Image Processing, and Graphics (NCVPRIPG) held at IIT Jodhpur on July 21–23, 2023. This challenge introduces a novel dataset comprising images of handwritten text contributed by 1,352 unique writers. Predominantly, these images feature handwritten Hindi text, but they also encompass a variety of elements such as numbers, mathematical symbols, and English text. Participants were tasked with developing a model capable of automatically determining whether the text in a given pair of images is authored by the same writer or different writers. The primary objective of this challenge was to advance research in the realm of handwritten text recognition. Throughout the competition, we registered 108 teams, with 18 teams submitting results for the validation dataset. Out of these, 13 teams provided submissions for the semi-final dataset. The top six teams from the semi-finals were subsequently invited to compete in the finals. Additional details about the challenge can be found at https://vl2g.github.io/challenges/wv2023/ .