Tavondo
HomeCompetitionsQuizGuidesPlatforms

No matching competitions

Try a platform, topic, or competition title

中文
HomeCompetitionsQuizGuidesPlatforms
中文RegisterSign in
Tavondo

Find AI contests that fit you—quiz, recommendations, catalog, and guides in one place.

Discover

CompetitionsQuizGuidesPlatforms

Account

Sign inRegisterSkill quiz

© 2026 Tavondo

中文
  1. Home
  2. Competitions
  3. PaddleOCR 算法模型挑战赛 - 赛题二:通用表格识别任务

Competition workspace

Baidu AI StudioComputer VisionIntermediateRegistration unverifiedRegistration unverified

PaddleOCR 算法模型挑战赛 - 赛题二:通用表格识别任务

适合已有 PaddlePaddle/PaddleOCR 使用经验、能训练和优化图像表格结构识别模型的团队。参赛者还需要能处理 PubTabNet 表格图像,并按指定 JSON 格式生成 HTML 结构预测;若希望进入决赛,还应能在 AtomGit 开源完整训练、推理代码和权重,并参加线下答辩。

Open official pageSign in to saveOpen guide

Suggested next step

Decide whether this contest fits your current stage before you sink time into the leaderboard.

Beginner score

55/100

Higher means a safer real starting point for your current stage.

Learning value

58/100

Measures whether the competition teaches habits you can transfer elsewhere.

Estimated effort

10-30 hours

Use this to check whether the task fits your current time budget.

Metric

Official scoring is published on the Ba…

If you cannot explain this metric clearly yet, you probably still need a bit of prep.

Who this competition fits

适合已有 PaddlePaddle/PaddleOCR 使用经验、能训练和优化图像表格结构识别模型的团队。参赛者还需要能处理 PubTabNet 表格图像,并按指定 JSON 格式生成 HTML 结构预测;若希望进入决赛,还应能在 AtomGit 开源完整训练、推理代码和权重,并参加线下答辩。

Official blurb (unedited)
PaddleOCR 旨在打造一套丰富、领先、且实用的 OCR 工具库,助力开发者训练出更好的模型,并应用落地。作为飞桨开源社区最热门的套件开源项目,PaddleOCR 拥有业界领先的 PP-OCR 模型和多场景通用的表格识别模型,广受众多开发者和用户的认可和使用。本次算法模型挑战赛聚焦通用 OCR 模型,要求开发者以当前飞桨开源模型为基线,不断突破技术瓶颈,研发出更加高效准确的新模型。 Official tags: 通用 OCR 模型. Sign-ups: 245.

Prep before joining

If the items below still feel unfamiliar, you usually get a better result by preparing first instead of rushing in.

PaddlePaddle

PaddleOCR

表格结构识别

PubTabNet 数据集

HTML 表格序列生成

JSONL 结果文件生成

Where the real difficulty shows up

The real friction is usually not library usage. It is validation, time allocation, and task framing.

核心难点是同时提高复杂表格结构的逐 Token 识别正确性并维持接近基线的推理速度:单个 Token 出错即会使该图像判错。提交不仅是跑出预测结果,还要求使用 PaddlePaddle 训练,并在复核时提供可复现最佳成绩的源码;进入决赛的队伍还需开源代码。

Read this first

This guide is the best pre-read if you want a cleaner start instead of trial-and-error.

A pre-flight checklist for image competitions when you only have free cloud GPU access.

Your First CV Baseline Checklist

Your decision

Use these fields to make a quick decision before you dive deeper.

Status
Registration unverified
Registration
Registration unverified
Eligibility
Needs official confirmation
Reward
¥100,000
Compute
Single GPU
Official metric
Official scoring is published on the Baidu AI Studio competition page.
Verification
Source page verified

Registration is unverified; confirm availability on the official page before investing.

Save to account

Sign in to save competitions and build your own shortlist.

Sign in to saveCreate account

Official source

Rules, files, submission details, and the live deadline still come from the official page.

Open official page

Related competitions

These competitions share a similar domain or difficulty level.

iFlytek Challenge

Computer Vision

高分辨率遥感影像变电站识别挑战赛

8d to register · Aug 10, 2026

iFlytek Challenge

Computer Vision

在线考试防作弊手机识别挑战赛

26d to register · Aug 28, 2026

iFlytek Challenge

Computer Vision

野生东北虎个体识别挑战赛

Registration deadline · Sep 3, 2026

iFlytek Challenge

Computer Vision

野生大熊猫个体识别挑战赛

Registration deadline · Sep 3, 2026

Data and review state

Separate what is confirmed from what still needs review. Official rules and deadlines win — report anything that looks wrong.

Source page verifiedMetric confirmedReviewed
Published on this site
Jul 21, 2026

Official metric is not published on the Baidu AI Studio competition listing; verify scoring on the competition page.