All Projects → fierceX → Dog-Breed-Identification-Gluon

fierceX / Dog-Breed-Identification-Gluon

Licence: other
Kaggle 120种狗分类,Gluon实现

Programming Languages

CSS
56736 projects
javascript
184084 projects - #8 most used programming language
python
139335 projects - #7 most used programming language
HTML
75241 projects

Projects that are alternatives of or similar to Dog-Breed-Identification-Gluon

Mxnet Im2rec tutorial
this simple tutorial will introduce how to use im2rec for mx.image.ImageIter , ImageDetIter and how to use im2rec for COCO DataSet
Stars: ✭ 97 (+115.56%)
Mutual labels:  mxnet, gluon
Mxnet.sharp
.NET Standard bindings for Apache MxNet with Imperative, Symbolic and Gluon Interface for developing, training and deploying Machine Learning models in C#. https://mxnet.tech-quantum.com/
Stars: ✭ 134 (+197.78%)
Mutual labels:  mxnet, gluon
D2l En
Interactive deep learning book with multi-framework code, math, and discussions. Adopted at 300 universities from 55 countries including Stanford, MIT, Harvard, and Cambridge.
Stars: ✭ 11,837 (+26204.44%)
Mutual labels:  mxnet, kaggle
Ko en neural machine translation
Korean English NMT(Neural Machine Translation) with Gluon
Stars: ✭ 55 (+22.22%)
Mutual labels:  mxnet, gluon
gluon-faster-rcnn
Faster R-CNN implementation with MXNet Gluon API
Stars: ✭ 31 (-31.11%)
Mutual labels:  mxnet, gluon
Gluon2pytorch
Gluon to PyTorch deep neural network model converter
Stars: ✭ 70 (+55.56%)
Mutual labels:  mxnet, gluon
Aws Machine Learning University Accelerated Nlp
Machine Learning University: Accelerated Natural Language Processing Class
Stars: ✭ 1,695 (+3666.67%)
Mutual labels:  mxnet, gluon
Sockeye
Sequence-to-sequence framework with a focus on Neural Machine Translation based on Apache MXNet
Stars: ✭ 990 (+2100%)
Mutual labels:  mxnet, gluon
Gluon Nlp
NLP made easy
Stars: ✭ 2,344 (+5108.89%)
Mutual labels:  mxnet, gluon
Imgclsmob
Sandbox for training deep learning networks
Stars: ✭ 2,405 (+5244.44%)
Mutual labels:  mxnet, gluon
Aws Machine Learning University Accelerated Cv
Machine Learning University: Accelerated Computer Vision Class
Stars: ✭ 1,068 (+2273.33%)
Mutual labels:  mxnet, gluon
ResidualAttentionNetwork
A Gluon implement of Residual Attention Network. Best acc on cifar10-97.78%.
Stars: ✭ 104 (+131.11%)
Mutual labels:  mxnet, gluon
Quantization.mxnet
Simulate quantization and quantization aware training for MXNet-Gluon models.
Stars: ✭ 42 (-6.67%)
Mutual labels:  mxnet, gluon
Mxnet Gluon Syncbn
MXNet Gluon Synchronized Batch Normalization Preview
Stars: ✭ 78 (+73.33%)
Mutual labels:  mxnet, gluon
Gluonrank
Ranking made easy
Stars: ✭ 39 (-13.33%)
Mutual labels:  mxnet, gluon
Mxnet Gluon Style Transfer
Neural Style and MSG-Net
Stars: ✭ 105 (+133.33%)
Mutual labels:  mxnet, gluon
Mxnet Centernet
Gluon implementation of "Objects as Points", aka "CenterNet"
Stars: ✭ 29 (-35.56%)
Mutual labels:  mxnet, gluon
Efficientnet
Gluon implementation of EfficientNet and EfficientNet-lite
Stars: ✭ 30 (-33.33%)
Mutual labels:  mxnet, gluon
Single Path One Shot Nas Mxnet
Single Path One-Shot NAS MXNet implementation with full training and searching pipeline. Support both Block and Channel Selection. Searched models better than the original paper are provided.
Stars: ✭ 136 (+202.22%)
Mutual labels:  mxnet, gluon
kaggle-dstl-satellite-imagery-feature-detection
6th place solution
Stars: ✭ 16 (-64.44%)
Mutual labels:  mxnet, kaggle

Dog-Breed-Identification-Gluon

Kaggle 120种狗分类比赛 Gluon实现代码
通过Gluon教程,边学边练。使用renset152_v1inception_v3进行模型融合得到0.27760分数

详细参见Gluon炼丹(Kaggle 120种狗分类,迁移学习加双模型融合)

环境

  • python 3.5+
  • mxnet 20171225以上
  • numpy 1.13.3
  • tqdm 4.11.2 (进度条)
  • matplotlib 2.0.2
  • tornado 4.5.2 (web展示网页的web框架和服务器,不需要可不装)

简易web网页展示

效果如图

训练好之后运行app.py即可 (需要安装tornado)

感悟

杨培文的代码中学到的,如果使用imagenet网络进行预训练,锁住特征层的话,那可以先把所有数据都都过一边特征层网络,这样在后面进行分类输出网络的训练时会省去很多时间和显存,并且在训练的时候可以把bs开到很大,而且在过特征层网络的时候,可以把bs设的小一点,而且不用反向传播,速度也不会太慢,这样就解决了在低配电脑上无法训练的问题。而且用最新版mxnet输出的特征数据非常小,70M左右,后面的分类层即使用cpu也能跑很顺畅

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].