From 3729c47136b950af6ae815fd8c1fc8fc035fb607 Mon Sep 17 00:00:00 2001 From: Ernestina <121707790@qq.com> Date: Thu, 10 Apr 2025 11:24:12 +0800 Subject: [PATCH 01/11] [WeeklyReport]2025.03.24~2025.04.06 --- .../[WeeklyReport]2025.03.24~2025.04.06.md | 29 +++++++++++++++++++ 1 file changed, 29 insertions(+) create mode 100644 WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.03.24~2025.04.06.md diff --git a/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.03.24~2025.04.06.md b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.03.24~2025.04.06.md new file mode 100644 index 00000000..d2eb50d6 --- /dev/null +++ b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.03.24~2025.04.06.md @@ -0,0 +1,29 @@ +### 姓名 + +邱文宇 + +### 实习项目 + +轻量高效表格识别新范式探索 + +### 本周工作 + +1. **调研行业内表格结构识别、框线补全、框线转excel/html的技术方案** + +2. **熟悉PaddleX表格识别项目** + + * 学习表格单元格检测模块、表格结构识别模块和自优化结果融合算法 + + * 熟悉工具代码 + +3. **问题疑惑与解答** + + * aistudio按照table_recognition_v2_tutorial.md教程进行数据集测评时报错? + + 答:导师的测试是正常的,排查代码版本和环境 + +### 下周工作 + +1. 进行框线补全实验 + +### 导师点评 From 3954d463cfcd477cb1424f51c2e43afa5086f192 Mon Sep 17 00:00:00 2001 From: Ernestina <121707790@qq.com> Date: Thu, 10 Apr 2025 18:06:12 +0800 Subject: [PATCH 02/11] Update [WeeklyReport]2025.03.24~2025.04.06.md --- .../ErnestinaQiu/[WeeklyReport]2025.03.24~2025.04.06.md | 2 -- 1 file changed, 2 deletions(-) diff --git a/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.03.24~2025.04.06.md b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.03.24~2025.04.06.md index d2eb50d6..1bbea98a 100644 --- a/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.03.24~2025.04.06.md +++ b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.03.24~2025.04.06.md @@ -25,5 +25,3 @@ ### 下周工作 1. 进行框线补全实验 - -### 导师点评 From 173c530432afb45c21f59c861c09af51dabb7d89 Mon Sep 17 00:00:00 2001 From: Ernestina <121707790@qq.com> Date: Sun, 27 Apr 2025 22:19:34 +0800 Subject: [PATCH 03/11] Create [WeeklyReport]2025.04.07~2025.04.27.md --- .../[WeeklyReport]2025.04.07~2025.04.27.md | 25 +++++++++++++++++++ 1 file changed, 25 insertions(+) create mode 100644 WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.07~2025.04.27.md diff --git a/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.07~2025.04.27.md b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.07~2025.04.27.md new file mode 100644 index 00000000..85c4c9d6 --- /dev/null +++ b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.07~2025.04.27.md @@ -0,0 +1,25 @@ +### 姓名 + +邱文宇 + +### 实习项目 + +轻量高效表格识别新范式探索 + +### 本周工作 + +1. **框线补全相关图神经网络算法调研** + +2. **框线补全实验** + + * 对框线不全数据集进行数据分析和挖掘 + + * 特征工程和设计子图聚类算法 + +3. **问题疑惑与解答** + + 暂无 + +### 下周工作 + +1. 继续进行框线补全实验 From b4ba8e131789d122a9c5dc15f49ba69dd0b9d9b6 Mon Sep 17 00:00:00 2001 From: Ernestina <121707790@qq.com> Date: Fri, 16 May 2025 23:16:43 +0800 Subject: [PATCH 04/11] Create [WeeklyReport]2025.04.28 - 2025.05.16.md [WeeklyReport]2025.04.28 - 2025.05.16.md --- .../[WeeklyReport]2025.04.28 - 2025.05.16.md | 27 +++++++++++++++++++ 1 file changed, 27 insertions(+) create mode 100644 WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.28 - 2025.05.16.md diff --git a/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.28 - 2025.05.16.md b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.28 - 2025.05.16.md new file mode 100644 index 00000000..92ea2415 --- /dev/null +++ b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.28 - 2025.05.16.md @@ -0,0 +1,27 @@ +### 姓名 + +邱文宇 + +### 实习项目 + +轻量高效表格识别新范式探索 + +### 本周工作 + +1. 梳理[TIES-2.0](https://github.com/shahrukhqasim/TIES-2.0)论文&代码和[caloGraphNN](https://github.com/jkiesele/caloGraphNN)并进行改造 + + TIES2.0项目已经长时间未维护,且存在大量未解决issues. + +2. **完成图像特征处理和图神经网络部分模型并根据TableMagic v2产线结构进行部分模型改造** + + + +3. **问题疑惑与解答** + + 暂无 + +### 下周工作 + +1. 完成TIES-2.0整体模型 + + From e27ef33b245bcb4b73321cf9c919bea19f21bd2d Mon Sep 17 00:00:00 2001 From: Ernestina <121707790@qq.com> Date: Fri, 16 May 2025 23:20:49 +0800 Subject: [PATCH 05/11] Update [WeeklyReport]2025.04.28 - 2025.05.16.md --- .../ErnestinaQiu/[WeeklyReport]2025.04.28 - 2025.05.16.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.28 - 2025.05.16.md b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.28 - 2025.05.16.md index 92ea2415..cf4237ae 100644 --- a/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.28 - 2025.05.16.md +++ b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.28 - 2025.05.16.md @@ -14,7 +14,7 @@ 2. **完成图像特征处理和图神经网络部分模型并根据TableMagic v2产线结构进行部分模型改造** - + 项目地址: [GitHub - ErnestinaQiu/PaddleX-TableRec at my-develop](https://github.com/ErnestinaQiu/PaddleX-TableRec/tree/my-develop) 3. **问题疑惑与解答** From 6c0109e39690e32287a5c225f0e41320f974eca2 Mon Sep 17 00:00:00 2001 From: Ernestina <121707790@qq.com> Date: Fri, 16 May 2025 23:24:22 +0800 Subject: [PATCH 06/11] Update [WeeklyReport]2025.04.28 - 2025.05.16.md --- .../ErnestinaQiu/[WeeklyReport]2025.04.28 - 2025.05.16.md | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.28 - 2025.05.16.md b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.28 - 2025.05.16.md index cf4237ae..55190267 100644 --- a/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.28 - 2025.05.16.md +++ b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.04.28 - 2025.05.16.md @@ -8,11 +8,11 @@ ### 本周工作 -1. 梳理[TIES-2.0](https://github.com/shahrukhqasim/TIES-2.0)论文&代码和[caloGraphNN](https://github.com/jkiesele/caloGraphNN)并进行改造 +1. 梳理[TIES-2.0](https://github.com/shahrukhqasim/TIES-2.0)论文&代码和[caloGraphNN](https://github.com/jkiesele/caloGraphNN)并进行优化 TIES2.0项目已经长时间未维护,且存在大量未解决issues. -2. **完成图像特征处理和图神经网络部分模型并根据TableMagic v2产线结构进行部分模型改造** +2. **完成图像特征处理和图神经网络部分模型并根据TableMagic v2产线结构进行部分模型优化** 项目地址: [GitHub - ErnestinaQiu/PaddleX-TableRec at my-develop](https://github.com/ErnestinaQiu/PaddleX-TableRec/tree/my-develop) @@ -23,5 +23,3 @@ ### 下周工作 1. 完成TIES-2.0整体模型 - - From 41ba1fd42dad9214cf942682271b809267e3ca33 Mon Sep 17 00:00:00 2001 From: Ernestina <121707790@qq.com> Date: Sun, 15 Jun 2025 23:28:15 +0800 Subject: [PATCH 07/11] Create [WeeklyReport]2025.05.16 - 2025.06.15.md --- .../[WeeklyReport]2025.05.16 - 2025.06.15.md | 21 +++++++++++++++++++ 1 file changed, 21 insertions(+) create mode 100644 WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.05.16 - 2025.06.15.md diff --git a/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.05.16 - 2025.06.15.md b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.05.16 - 2025.06.15.md new file mode 100644 index 00000000..e6e49686 --- /dev/null +++ b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.05.16 - 2025.06.15.md @@ -0,0 +1,21 @@ +### 姓名 + +邱文宇 + +### 实习项目 + +轻量高效表格识别新范式探索 + +### 本周工作 + +1. 表格单元格及行列结构识别模型优化和制作框线补全程序 + + 项目地址: [GitHub - ErnestinaQiu/PaddleX-TableRec at my-develop](https://github.com/ErnestinaQiu/PaddleX-TableRec/tree/my-develop) + +2. **问题疑惑与解答** + + 暂无 + +### 下周工作 + +1. 表格单元格及行列结构识别模型优化和框线补全程序指标优化 From 1628c6d66679dc826b14e119aaad65d6540585a1 Mon Sep 17 00:00:00 2001 From: Ernestina <121707790@qq.com> Date: Sun, 3 Aug 2025 21:57:54 +0800 Subject: [PATCH 08/11] Create [WeeklyReport]2025.07.18 - 2025.08.01.md --- .../[WeeklyReport]2025.07.18 - 2025.08.01.md | 23 +++++++++++++++++++ 1 file changed, 23 insertions(+) create mode 100644 WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.07.18 - 2025.08.01.md diff --git a/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.07.18 - 2025.08.01.md b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.07.18 - 2025.08.01.md new file mode 100644 index 00000000..bca79611 --- /dev/null +++ b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.07.18 - 2025.08.01.md @@ -0,0 +1,23 @@ +### 姓名 + +邱文宇 + +### 实习项目 + +轻量高效表格识别新范式探索 + +### 本周工作 + +1. 数据集收集、制作和benchmark确认 + +2. 目标检测和实例分割方案分析和实验 + + + +1. **问题疑惑与解答** + + 暂无 + +### 下周工作 + +1. 实验分割方案构建和测试 From d2f73aa018d6aafbaab7cb729f0944bb9a243ced Mon Sep 17 00:00:00 2001 From: Ernestina <121707790@qq.com> Date: Sun, 21 Sep 2025 22:27:26 +0800 Subject: [PATCH 09/11] Create [WeeklyReport]2025.09.11 - 2025.09.21.md --- .../[WeeklyReport]2025.09.11 - 2025.09.21.md | 22 +++++++++++++++++++ 1 file changed, 22 insertions(+) create mode 100644 WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.09.11 - 2025.09.21.md diff --git a/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.09.11 - 2025.09.21.md b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.09.11 - 2025.09.21.md new file mode 100644 index 00000000..70045e87 --- /dev/null +++ b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.09.11 - 2025.09.21.md @@ -0,0 +1,22 @@ +### 姓名 + +邱文宇 + +### 实习项目 + +轻量高效表格识别新范式探索 + +### 本周工作 + +1. 对v2在测试集上的指标表现进行归因分析 + +1. **问题疑惑与解答** + + * 对v2在新数据集上的表现以及所适配的场景进行了讨论 + + 解答:明确了目前v2模型的适配场景和基于当前测试集的痛点,确定了优先基于v2框架进行部分算法模块进行优化的思路 + +### 下周工作 + +1. v2在测试集上的指标表现进行归因分析 +2. 明确优化思路和具体算法 From 8951c451b4c33ee9a8fd379b219412a6d029ea18 Mon Sep 17 00:00:00 2001 From: Ernestina <121707790@qq.com> Date: Sun, 19 Oct 2025 23:40:42 +0800 Subject: [PATCH 10/11] Create [WeeklyReport]2025.09.22 - 2025.10.19.md --- .../[WeeklyReport]2025.09.22 - 2025.10.19.md | 15 +++++++++++++++ 1 file changed, 15 insertions(+) create mode 100644 WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.09.22 - 2025.10.19.md diff --git a/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.09.22 - 2025.10.19.md b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.09.22 - 2025.10.19.md new file mode 100644 index 00000000..0fe6f8c2 --- /dev/null +++ b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.09.22 - 2025.10.19.md @@ -0,0 +1,15 @@ +### 姓名 + +邱文宇 + +### 实习项目 + +轻量高效表格识别新范式探索 + +### 本周工作 + +1. 对v2在测试集上的指标表现进行归因分析 + +### 下周工作 + +1. 大类问题定位与优化实验 From a66d712a56c462091053d6aa8629611cd975c372 Mon Sep 17 00:00:00 2001 From: Ernestina <121707790@qq.com> Date: Sun, 23 Nov 2025 15:14:09 +0800 Subject: [PATCH 11/11] Create [WeeklyReport]2025.11.10 - 2025.11.23.md --- .../[WeeklyReport]2025.11.10 - 2025.11.23.md | 15 +++++++++++++++ 1 file changed, 15 insertions(+) create mode 100644 WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.11.10 - 2025.11.23.md diff --git a/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.11.10 - 2025.11.23.md b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.11.10 - 2025.11.23.md new file mode 100644 index 00000000..ac4286bd --- /dev/null +++ b/WeeklyReports/Hackathon_8th/ErnestinaQiu/[WeeklyReport]2025.11.10 - 2025.11.23.md @@ -0,0 +1,15 @@ +### 姓名 + +邱文宇 + +### 实习项目 + +轻量高效表格识别新范式探索 + +### 本周工作 + +1. 大类问题定位与优化实验 + +### 下周工作 + +1. 大类问题定位与优化实验