From 1fb7db6ec36b769bc139f506d2d569be71502c75 Mon Sep 17 00:00:00 2001 From: Spacesick Date: Mon, 12 Jun 2023 18:12:45 +0700 Subject: [PATCH] Update README.md --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 461c60f..90df681 100644 --- a/README.md +++ b/README.md @@ -28,7 +28,7 @@ We collected our Dataset various wound type from kaggle. We got 431 wound type i ### Architecture -We use the technique of transfer learning by adding multiple layers of fully connected networks on top of an InceptionV3 model pre-trained on the imagenet dataset. We freeze all the layers in the pre-trained InceptionV3 except for the last 12. We saved our best performing model in .h5 format and its weights in a .ckpt file, which so far has reached a validation accuracy of 0.8315. The latest checkpoints can be found [here](https://drive.google.com/drive/folders/1JRxoLPASzwMdFGjUG7gDhSLfg6zoRx-t?usp=sharing). History of build/develop model before fix model can be found [here](https://github.com/dhuhaardha/Model_Capstone) +We use the technique of transfer learning by adding multiple layers of fully connected networks on top of an InceptionV3 model pre-trained on the imagenet dataset. We freeze all the layers in the pre-trained InceptionV3 except for the last 12. We saved our best performing model in .h5 format and its weights in a .ckpt file, which so far has reached a validation accuracy of over 0.83 for the injury classification task and over 0.89 for the accident classification task. The dataset, latest checkpoints, and various model formats can be found [here](https://drive.google.com/drive/folders/1JRxoLPASzwMdFGjUG7gDhSLfg6zoRx-t?usp=sharing). Legacy versions of the model can be found [here](https://github.com/dhuhaardha/Model_Capstone). ## Cloud computing and deployment