From 03fb04fc1414010dc8adb0b068528c702397c14f Mon Sep 17 00:00:00 2001 From: littletomatodonkey <2120160898@bit.edu.cn> Date: Mon, 21 Jun 2021 17:12:49 +0800 Subject: [PATCH] fix readme (#945) --- README.md | 149 +----------------- README_ch.md | 148 +++++++++++++++++ README_en.md | 6 +- docs/images/mainpage/cartoon_vis_en_v1.0.gif | Bin 0 -> 962355 bytes docs/images/mainpage/cartoon_vis_v1.0.gif | Bin 0 -> 924600 bytes docs/images/mainpage/logo_vis_en_v1.0.gif | Bin 0 -> 944352 bytes docs/images/mainpage/logo_vis_v1.0.gif | Bin 0 -> 1006154 bytes docs/images/mainpage/product_vis_en_v1.0.gif | Bin 0 -> 1096833 bytes docs/images/mainpage/product_vis_v1.0.gif | Bin 0 -> 1182769 bytes .../mainpage/recognition_pipeline_en.png | Bin 0 -> 155908 bytes docs/images/mainpage/vehicle_vis_en_v1.0.gif | Bin 0 -> 1111379 bytes 11 files changed, 152 insertions(+), 151 deletions(-) mode change 100644 => 120000 README.md create mode 100644 README_ch.md create mode 100644 docs/images/mainpage/cartoon_vis_en_v1.0.gif create mode 100644 docs/images/mainpage/cartoon_vis_v1.0.gif create mode 100644 docs/images/mainpage/logo_vis_en_v1.0.gif create mode 100644 docs/images/mainpage/logo_vis_v1.0.gif create mode 100644 docs/images/mainpage/product_vis_en_v1.0.gif create mode 100644 docs/images/mainpage/product_vis_v1.0.gif create mode 100644 docs/images/mainpage/recognition_pipeline_en.png create mode 100644 docs/images/mainpage/vehicle_vis_en_v1.0.gif diff --git a/README.md b/README.md deleted file mode 100644 index ba4189f0..00000000 --- a/README.md +++ /dev/null @@ -1,148 +0,0 @@ -简体中文 | [English](README_en.md) - -# PaddleClas - -## 简介 - -飞桨图像识别套件PaddleClas是飞桨为工业界和学术界所准备的一个图像识别任务的工具集,助力使用者训练出更好的视觉模型和应用落地。 - -**近期更新** - -- 2021.06.22,23,24 PaddleClas官方研发团队带来技术深入解读三日直播课,6月22日、23日、24日晚上20:30,[直播地址](https://live.bilibili.com/21689802) -- 🔥🔥🔥: 2021.06.16 PaddleClas v2.2版本升级,集成Metric learning,向量检索等组件。新增商品识别、动漫人物识别、车辆识别和logo识别等4个图像识别应用。新增LeViT、Twins、TNT、DLA、HarDNet、RedNet系列30个预训练模型。 -- 2021.05.14 添加`SwinTransformer` 系列模型。 -- 2021.04.15 添加`MixNet_L`和`ReXNet_3_0`系列模型。 - -[more](./docs/zh_CN/update_history.md) - -## 特性 - -- 实用的图像识别系统:集成了目标检测、特征学习、图像检索等模块,广泛适用于各类图像识别任务。 -提供商品识别、车辆识别、logo识别和动漫人物识别等4个场景应用示例。 - -- 丰富的预训练模型库:提供了35个系列共164个ImageNet预训练模型,其中6个精选系列模型支持结构快速修改。 - -- 全面易用的特征学习组件:集成arcmargin, triplet loss等12度量学习方法,通过配置文件即可随意组合切换。 - -- SSLD知识蒸馏:14个分类预训练模型,精度普遍提升3%以上;其中ResNet50_vd模型在ImageNet-1k数据集上的Top-1精度达到了84.0%, -Res2Net200_vd预训练模型Top-1精度高达85.1%。 - -- 数据增广:支持AutoAugment、Cutout、Cutmix等8种数据增广算法详细介绍、代码复现和在统一实验环境下的效果评估。 - -## 图像识别系统效果展示 -
ZvEdE5(
zW&(7pV9>ax!svCjq^-@IVrsir%t(2*FBQ`Nj$}QK+G3iUi6Q>K{150bS|*wYN=OWp
zNbgz+qqeasXzSK%#DO;IZrfj%_s2df4i{N1yfyI@b;{&&UnTwB9!`nClB1zyiO1nB
zr|UEMm^9MKT{rxjFcx7UUDWu47e!l1OQ|d*0g>pd4<;2K5mhfm&H=Y}YDpR{pJKU?
ztf^zWspDj!5P)g|`j>l!S_sT;u!4X)kAMp0=z@^rf=Gp+=v@%TbC6-bn&GIfF8|6p
z8;HpvL0?ywaUEo%s%ImnXY0CZw*%pD0NM6K(1k#BU3xCFt1gj1x_v#@zj`k1t8SEQ
z&iz0}R1l-WnzyR%XT07q3hYG31%}o
za=hl(J-Qlx7GoiYq&yVl;!?IJyKr8!628&$j={iCA;oupXlKDWIAW~t;g5-0PJ`F4
zR;R$(9p*%{KiEDTfny@{tD<^A#Z8B)4LA9~#(oKJZtJVUwk@>|`N+Bj%eYr+Ri4#R
zWnSX#WVSkX7JZt?%P%I?0T%gy^+ j9ahI#0pI>s|jyJm&!hpF0xRF|RE?xAm_KL4p8te2OdYrzp_V3>c)vL}4|16jJ{
zzAs2IzZmnW*Slj)(&?17rP*4!5$#NX%gf()VTR ZsuwyqK{!(!;bPK3OHyy;6)?3q5mt~BsSs|j@P_r``)9&g91q)UIt)y57XY7+3
zc 0V5qjNtJ!o`#onZgVsh=^b<5!e
zFHnksTZ7O2N{4UVV*nhjy2-+P<%j1g27DrA(-kxBH)ns`VGxVJ1XX7b1{99Q4njlm
zyq(vNsw>Zb;8o=S)spHuD1|QxPvblkJa&te)|kA4IKFi%9GDVbebS)~7uC*C*Ab*?
z+^xmy>TjqMH36DwGl?8n7^g{lrCQ%RjXsu^sfvMOhL?!OXk;hpiFCA9jG4?I6QvR$
z8Ew~V<&r@b)Fo+ivHYomHxJ#n%NIS)cqW#nTShCEKjxk%q#ZP}8o&h;FYER%TZ|)-
z17^BcH2-3Go@dYo)6#HaL>uV1@*?4eU3%#xy~ IW|hqlwHS?DjODH<$haf6vP+QC<4q`5rWt6!R*|7858b;iB{
z=W6N&J3FPnTY!`nx9B1yxE}A3jWj0os1s*j1ZuVIj+Z;Bv=}!wXwufsJ;d8%Tfs7b
z_Dnae=60u|F#{e5!I#w
z@G?9(d`H<_ix6IEM^V)mUvctK56D7Q!Hl{BHJ|~-C!&Ae_S;X}7*ZD>C$d^ImEs`|
zdr4{B5j>>l|+Po`cY7t2A8+a7?4U!eYu!hJ@XE1)wy5yJMpjR#DfN9ftISz-#5;
z!!h~yY>~zhPBet66Hwaf>Q;gnR)1flzDk$xO1)*m$*94KOjA8#Bs`w8F8evi_^&Qt
z=|XckQ9@?Mu6YN<(@C0jqK({@=JBT6`lQ?;Ec0E_zB1ru@ud@Oow`gIH~-6p8UV8L
zG(RovO}9eHalrzuv;09KFn&JsF0jtI3xty%z)rZhskH*bRF8hn0lJEq#1fEn{Cn2(
z8ItqgTQQSagTDd+yx%Y^TR;|E#TPs%LdlW$1bE`n0CQb-rdc>PQHU ~B<~FgceyvBd0dKLw`l;l4(<
zO{8_8TrOJgSTnKB`JHBoEw!=37$G^ca@pfx=|~{dS>`tj`LtJ6b$sI}Tp6pHYC`mw
zu%SO?smMLM>)T`Vi6f&v6COar)jz@G*%p7j3gF`ol)Yu(?6DmpWq5rO!DO}B@}hrH
zAv4(gpV|W+2vY8M_iU;7=Z3PLA3iy$Nl}ykjrW-cO2}zUN8X}t?rnzGMH^$yU&p=+
zp8HE3{%|h~PLAIzH@WBWFfZiyH=XGsy2;DA3Y4*( Xov
zX61lT3IGJ6USk=vTJwZ6@Jfg1wKENm(ut$%0aR-vNiY`I+z79%ktEhItv1I7Vl^GD
zK)9kKRprIH0V_I=_S-pg^;o&Gb_rq*{aU+Dorh~%t}@|OBMqCHUjUW}rlu*$ALHjA
zqG4iF>^;N-XTfh>^sd0Z6-Qj>YaH0$`UIgblyS|?
z~#mcNHhD{0RFH{Szpi5Qq~7ygC6
D_z~jie
fvefNm+WvS#uPHjGN*Xh@duy
GNXmBk#l^GweG5AmRY73r@t&C=)JXEh<&
zse6@)goRcmT^{X$T-7SkyVY8_s>VJbFH&;6Z;40!d6CATp!yKUtLOf4P!Vl+7X&Pc
za^Lmyfz6J#U+<0YLO
S4G>9{;8q_pwoOwZ*<81-S?43dpnE6QWSW*^2%$$Dp5oYcMfLdiU
zRjjm#DoOLG{<4+QKUfB6&71mTT7W3jr1~zAnm1vG!3l
zL`-CBu;dVQMbl5?oOF{juH|sg??5}N&5@umjA0lhk2}XxKB`Llb(nzEnf&EMwX+!6
zS*-;kMgj>0fxIJFcfB#`ha^8S3d`Bx?Ep#Iw*O2%NfAJsrstVrvR1k*wh#3o3u
z4mf52X93(vjj%)pi6C1dIPyfIIPO3f-LgWIVI3+S-TKVl%2tB_>J|VCA<4VISI7H4
zp2(`dg%;Rk{gfXlXBoW(gWTL?Gv8x7UrNrZ(blr!I9@QFV)a>W<$b%YS@%mK0E%j?
z>TtC$w>#JAFa@FBEpH_xUL~2S=qets*Qz4}<|9Pt0baKd@Ss2Pt
z9W9t2hwyyPA)YEwSyeyc$+gfp58ZOe5~Bomih@3|L+GxoETp3JO%T