upread README

README.md

@@ -8,13 +8,18 @@ We design the UF (Unweighted Fusion) block and CWF (Complementary Weighted Fusio
 
 Decription of files
 -----
->lanms/                      A C++ version of NMS
->nets/
->>resnet_utils.py            Contains building blocks for various versions of Residual Networks
->>resnet_v1.py               Resnet V1 model implemented with [Slim](https://github.com/tensorflow/models/tree/master/research/slim)
->>vgg.py                     VGG model implemented with [Slim](https://github.com/tensorflow/models/tree/master/research/slim)
->data_util.py                A base data generator
->image_augmentation.py       Various image augmentation methods
->multigpu_train
->eval_all_ckpt_*.py          Evaluate the correspoding models
->
+>lanms/                      A C++ version of NMS <br>
+>nets/<br>
+>>resnet_utils.py            Contains building blocks for various versions of Residual Networks<br>
+>>resnet_v1.py               Resnet V1 model implemented with [Slim](https://github.com/tensorflow/models/tree/master/research/slim)<br>
+>>vgg.py                     VGG model implemented with [Slim](https://github.com/tensorflow/models/tree/master/research/slim)<br>
+>data_util.py                A base data generator<br>
+>oxford_R01.py               Data processor for Oxford dataset<br>
+>VIVA_R01.py                 Data processor for VIVA dataset<br>
+>image_augmentation.py       Various image augmentation methods<br>
+>multigpu_train_*.py         Train models<br>
+>eval_all_ckpt_*.py          Evaluate the correspoding models<br>
+>resnet_v1_model_*.py        SIFCN with Resnet V1 50 as the backbone network<br>
+>vgg16_model_*.py            SIFCN with VGG16 as the backbone network<br>
+>*_multi*.py                 The multi-scale loss discribed in the paper is used<br>
+>*_weighted_fusion*.py       The CWF block is used<br>, if not marked, the UF block is used as default<br>
\ No newline at end of file