From 73825d9017491270ed66c4ae8a041a251f89669c Mon Sep 17 00:00:00 2001 From: heitorrapela Date: Fri, 18 May 2018 20:41:18 -0300 Subject: [PATCH] Update MLP value of Fashion-Mnist --- README.ja.md | 1 + README.md | 1 + README.zh-CN.md | 1 + 3 files changed, 3 insertions(+) diff --git a/README.ja.md b/README.ja.md index 5346506..b8a3034 100644 --- a/README.ja.md +++ b/README.ja.md @@ -171,6 +171,7 @@ scikit-learn ベースの自動ベンチマーキング・システムを構築 |AlexNet with Triplet loss| None | 0.899 | - | [@Cenk Bircanoğlu](https://github.com/cenkbircanoglu) | [:link:](https://github.com/cenkbircanoglu/openface/tree/master/fashion_mnist)| |SqueezeNet with cyclical learning rate 200 epochs| None| 0.900| - | [@snakers4](https://github.com/snakers4) | [:link:](https://github.com/zalandoresearch/fashion-mnist/files/1263340/squeeze_net_mnist.zip)| |Dual path network with wide resnet 28-10|standard preprocessing (mean/std subtraction/division) and augmentation (random crops/horizontal flips) |0.957|-|[@Queequeg](https://github.com/Queequeg92)|[:link:](https://github.com/Queequeg92/DualPathNet)| +|MLP 256-128-100| None | 0.8833| - | [@heitorrapela](https://github.com/heitorrapela)| [:link:](https://github.com/heitorrapela/fashion-mnist-mlp)| |VGG16 26M parameters | None | 0.935| - | [@QuantumLiu](https://github.com/QuantumLiu)|[:link:](https://github.com/QuantumLiu/fashion-mnist-demo-by-Keras) [:link:](https://zhuanlan.zhihu.com/p/28968219)| |WRN-28-10| standard preprocessing (mean/std subtraction/division) and augmentation (random crops/horizontal flips) | 0.959 | -| [@zhunzhong07](https://github.com/zhunzhong07)|[:link:](https://github.com/zhunzhong07/Random-Erasing)| |WRN-28-10 + Random Erasing| standard preprocessing (mean/std subtraction/division) and augmentation (random crops/horizontal flips) | 0.963 | -| [@zhunzhong07](https://github.com/zhunzhong07)|[:link:](https://github.com/zhunzhong07/Random-Erasing)| diff --git a/README.md b/README.md index 022dfc5..6d0775a 100644 --- a/README.md +++ b/README.md @@ -171,6 +171,7 @@ The table below collects the submitted benchmarks. Note that **we haven't yet te |AlexNet with Triplet loss| None | 0.899 | - | [@Cenk Bircanoğlu](https://github.com/cenkbircanoglu) | [:link:](https://github.com/cenkbircanoglu/openface/tree/master/fashion_mnist)| |SqueezeNet with cyclical learning rate 200 epochs| None| 0.900| - | [@snakers4](https://github.com/snakers4) | [:link:](https://github.com/zalandoresearch/fashion-mnist/files/1263340/squeeze_net_mnist.zip)| |Dual path network with wide resnet 28-10|standard preprocessing (mean/std subtraction/division) and augmentation (random crops/horizontal flips) |0.957|-|[@Queequeg](https://github.com/Queequeg92)|[:link:](https://github.com/Queequeg92/DualPathNet)| +|MLP 256-128-100| None | 0.8833| - | [@heitorrapela](https://github.com/heitorrapela)| [:link:](https://github.com/heitorrapela/fashion-mnist-mlp)| |VGG16 26M parameters | None | 0.935| - | [@QuantumLiu](https://github.com/QuantumLiu)|[:link:](https://github.com/QuantumLiu/fashion-mnist-demo-by-Keras) [:link:](https://zhuanlan.zhihu.com/p/28968219)| |WRN-28-10| standard preprocessing (mean/std subtraction/division) and augmentation (random crops/horizontal flips) | 0.959 | -| [@zhunzhong07](https://github.com/zhunzhong07)|[:link:](https://github.com/zhunzhong07/Random-Erasing)| |WRN-28-10 + Random Erasing| standard preprocessing (mean/std subtraction/division) and augmentation (random crops/horizontal flips) | 0.963 | -| [@zhunzhong07](https://github.com/zhunzhong07)|[:link:](https://github.com/zhunzhong07/Random-Erasing)| diff --git a/README.zh-CN.md b/README.zh-CN.md index ab37aaa..1749561 100644 --- a/README.zh-CN.md +++ b/README.zh-CN.md @@ -175,6 +175,7 @@ data = input_data.read_data_sets('data/fashion', source_url='http://fashion-mnis |AlexNet with Triplet loss| None | 0.899 | - | [@Cenk Bircanoğlu](https://github.com/cenkbircanoglu) | [:link:](https://github.com/cenkbircanoglu/openface/tree/master/fashion_mnist)| |SqueezeNet with cyclical learning rate 200 epochs| None| 0.900| - | [@snakers4](https://github.com/snakers4) | [:link:](https://github.com/zalandoresearch/fashion-mnist/files/1263340/squeeze_net_mnist.zip)| |Dual path network with wide resnet 28-10|standard preprocessing (mean/std subtraction/division) and augmentation (random crops/horizontal flips) |0.957|-|[@Queequeg](https://github.com/Queequeg92)|[:link:](https://github.com/Queequeg92/DualPathNet)| +|MLP 256-128-100| None | 0.8833| - | [@heitorrapela](https://github.com/heitorrapela)| [:link:](https://github.com/heitorrapela/fashion-mnist-mlp)| |VGG16 26M parameters | None | 0.935| - | [@QuantumLiu](https://github.com/QuantumLiu)|[:link:](https://github.com/QuantumLiu/fashion-mnist-demo-by-Keras) [:link:](https://zhuanlan.zhihu.com/p/28968219)| |WRN-28-10| standard preprocessing (mean/std subtraction/division) and augmentation (random crops/horizontal flips) | 0.959 | -| [@zhunzhong07](https://github.com/zhunzhong07)|[:link:](https://github.com/zhunzhong07/Random-Erasing)| |WRN-28-10 + Random Erasing| standard preprocessing (mean/std subtraction/division) and augmentation (random crops/horizontal flips) | 0.963 | -| [@zhunzhong07](https://github.com/zhunzhong07)|[:link:](https://github.com/zhunzhong07/Random-Erasing)|