Discussion for April Plan

[innerlee]

Here is a list of candidate topics that we could focus on in April. We will discuss and select a portion of them in a meeting later on. Feel free to add or comment on :)

Installation

• [x] Compilation-free: Move cpps to mmcv @jeffreykuang
• [x] Simplify installation instructions @innerlee
• [x] Release to pypi (wait for unittest) @jeffreykuang

Documentation

• [ ] Tutorials on fine-tuning/training on user's data
• [x] Chinese documentation @quincylin1

Demo

• [ ] Documentation for demos
• [x] End2end demo for text detection & recognition @cuhk-hbsun
• [x] End2end demo of KIE @innerlee
• [x] Colab @quincylin1
• [x] Online demo @yuexy
• [x] Add demo cfg @innerlee @cuhk-hbsun

Auxiliary Tools

• [ ] Log analysis
• [ ] Model complexity
• [ ] onnx
• [ ] config printer
• [ ] model publishing

Benchmark

• [ ] benchmarks with paddleocr, chineseocr, easyocr, chineseocr_lite
  • [ ] Speed
  • [ ] Accuracy
  • [x] Algorithm coverage @jeffreykuang
  • [x] Feature set @jeffreykuang

Multilingual

• [ ] General discussion

Datasets

• [ ] Improve preparation scripts & documents

More Algorithms

• [x] NER @huaqiang
• [x] FCENet @HolyCrap96
• [x] DRRG @HolyCrap96

[2793145003]

Any plan about SATRN? (On Recognizing Texts of Arbitrary Shapes with 2D Self-Attention, https://github.com/Media-Smart/vedastr)

[jeffreykuang]

Any plan about SATRN? (On Recognizing Texts of Arbitrary Shapes with 2D Self-Attention, https://github.com/Media-Smart/vedastr)

If you are interested in reimplementing it, welcome to join us!

[innerlee]

@2793145003 Thanks for the propasal, we will consider it in later iterations :D

[priyanshu-sinha-trysolvio]

@innerlee @cuhk-hbsun Did you add the onnx support? as it is not visible in the iteration plans of May-June.

[innerlee]

Not yet. Please raise/up-vote this feature at https://github.com/open-mmlab/mmocr/issues/39 to help us prioritize

[priyanshu-sinha-trysolvio]

Done.