Update README.md

Author: paarthneekhara

README.md

@@ -48,6 +48,22 @@ python extract_fc7.py --split=val
 - Answers are mapped to 1000 word vocabulary, covering 87% answers across training and validation datasets.
 - The LSTM+VIS model is defined in vis_lstm.py. The input tensors for training are fc7 features, Questions(Word indices upto 22 words), Answers(one hot encoding vector of size 1000). The model depicted in the figure is implemented with 2 LSTM layers by default(num_layers in configurable).
 
+## Sample Predictions
+
+The fun part! Try it for yourself. Make sure you have tensorflow installed. Download the data files/trained model from [this link][9] and save them in the ```Data/``` directory. Test for an image using:
+```
+python predict.py --image_path="Data/sample.jpg" --question="Which animal is this?" --model_path="Data/model7.ckpt"
+```
+| Image        | Question           | Top Answers (left to right)  |
+| ------------- |:-------------:| -----:|
+| ![](http://i.imgur.com/j4FiEaS.jpg)      | What color is the signal? | red, green, yellow|
+| ![](http://i.imgur.com/FUR7k0y.jpg)      | What animal is this? | giraffe, cow, horse|
+| ![](http://i.imgur.com/VrGUves.jpg)      | What animal is this? | cat, dog, giraffe|
+| ![](http://i.imgur.com/yk53y1Y.jpg)      | What color is the frisbee that is in the dog's mouth? | white, brown, red|
+| ![](http://i.imgur.com/yk53y1Y.jpg)      | What color is the frisbee that is upside down? | red, white, blue|
+| ![](http://i.imgur.com/ifcccpd.jpg)      | What are they playing with? | frisbee, soccer ball, soccer|
+| ![](http://i.imgur.com/VrjUbgH.jpg)      | What is in the standing person's hand? | bat, glove, ball|
+
 ## References
 - [Exploring Models and Data for Image Question Answering][1]
 - [Torch implementation of VQA][2]
@@ -61,3 +77,4 @@ python extract_fc7.py --split=val
 [6]: http://visualqa.org/
 [7]: https://github.com/ry/tensorflow-vgg16
 [8]: https://github.com/jazzsaxmafia/show_attend_and_tell.tensorflow
+[9]: https://github.com/ry/tensorflow-vgg16