Update README.md

README.md

@@ -1,2 +1,22 @@
 # DeepCar
 Fine-grained detection on Vehicle Model/Make
+
+## Dataset
+Training dataset consisted of 163/1,716 vehicle make/models from [CompCars dataset](http://mmlab.ie.cuhk.edu.hk/datasets/comp_cars/index.html)[1]
+
+
+## Fine-tune VGG
+#### Architecture
+A [VGG16](http://arxiv.org/pdf/1409.1556.pdf) model pre-trained on ImageNet was fine-tuned with CompCars dataset (16970/776 train/valid images - 115 vehicles/classes)
+
+#### Results
+Accuracy: **93.2% top-5** in **200 epochs**
+Base learning rate of 0.001 and batch size of 64 were used. 
+
+## RA-CNN Look closer to see better
+#### Architecture
+#### Results
+
+
+### References
+[1] Linjie Yang, Ping Luo, Chen Change Loy, Xiaoou Tang. A Large-Scale Car Dataset for Fine-Grained Categorization and Verification, In Computer Vision and Pattern Recognition (CVPR), 2015.