OSDN

Keras implementation for the research paper "Towards Open Set Deep Networks" A Bendale, T Boult, CVPR 2016

Original Implementation: https://github.com/abhijitbendale/OSDN

This repo has Keras wrapper for the above research paper. Full code plus ipython notebook is also avaliable.

jupyter notebook Softmax.ipynb

or

open notebook with nbviewer by clicking on this link

https://nbviewer.jupyter.org/github/aadeshnpn/OSDN/blob/master/Softmax.ipynb

If you have any question feel free to create an issue.

How to run the existing code

• Step 1: Train a CNN model for the dataset you choice
• Step 2: Load the trained model
• Step 3: Load the training data you trained the DNN model
• Step 4: Create a mean activation vector (MAV) and perform weibull fit model
• Step 5: Pass the sample to compute openmax and evaluate the output from openmax, original label, and softmax
• Step 6: Test the trained openmax to images from different distribution

Refer to the main.py for detail implementation