* Add S2ANet model for Oriented Object Detection.

* add docs for custom dataset, reader and model technical

* modify docs according to review, test=document_fix

* modify READER.md, test=document_fix

* add config annotation doc, test=document_fix

* update doc of config

* mlp_dim-->out_channel of TwoFCHead

* update doc, test=document_fix

* update doc, test=document_fix

* update quick_start

* update get_started

* update doc, test=document_fix

* Update install doc and deploy doc.

* fix dide link, test=document

* fix dide link, test=document

* support mask_rcnn for kunlun

* support ppyolo for kunlun

* Update QUICK_STARTED_cn.md (#1757)

fix typo

* cherry-pick #1757 and #1473, test=document_fix

* update v0.5 release note

* update some comments

* Add PrepareData-->SelectModel-->Train-->Eval-->Test-->Infer-->Slim-->Deploy document DetectionPipeline.md, test=document_fix

* enhance voc to coco docs

* fix voc annotation float type

* add PrepareDataSet.md, test=document_fix

* update PrepareDataSet, test=document_fix

* update PrepareDataSet, test=document_fix

Repalce fruit dataset with roadsign dataset, and update QUICK_STARTED, use PP-YOLO as demo, test=document_fix. (#1407)

* update INSTALL

* fix typo

* add demo image

* add use_gpu annotation

* fix link

* update style

* update style

* add install paddle

* use python3

* update INSTALL_cn.md

* update INSTALL_cn.md

* update INSTALL_cn.md

* fix typo

add fleet training

* fix voc dataset and docs

* add custom datsset demo

Add image for visualdl board.

* fix some docs link error

* rename config_doc

* fix export_model doc link error

* sort out release/0.3 docs

* add new CHANGELOG

* add yolov4 in changelog

* Update docs
* Update README.md

* optimize python path

* Remove `pycocotools` from `requirements.txt`

git version is required for numpy > 1.18

* Remove `cython`, add pip install doc

* add config link in model zoo

* fix link error

* add anchor free