CV_backend

Machine Learning backend to detect fake images

• Setup Azure instance with GPU (N6)
• Install NVIDIA driver and CUDA
• Build OpenCV3
• Install Deep Learning libraries (Chainer, Darknet, Tensorflow)
• Build Google reverse search hack
• Build ELA algorithm with OpenCV
• Implement ELA image fakeness scoring
• Use Darknet for object detection
• Use Chainer for caption generation
• Setup flask API to handle requests
• Train neural network to recognize weather, number plate, buildings, signs
• Compare analysis results with original Tweet data
• Return JSON report data

Azure instance

• Standard NC6 (6 cores, 56 GB memory)
• (West US 2)
• Linux Ubuntu 16.04

NVIDIA CUDA 8.0 for Ubuntu 16.04

https://developer.nvidia.com/compute/cuda/8.0/Prod2/local_installers/cuda_8.0.61_375.26_linux-run

CUDNN 5.1

https://developer.nvidia.com/compute/machine-learning/cudnn/secure/v5.1/prod_20161129/8.0/cudnn-8.0-linux-x64-v5.1-tgz

Env variable

Add to .bashrc:

#Define CUDA_HOME environment variable
export CUDA_HOME=/usr/local/cuda-8.0
#Define LD_LIBRARY_PATH environment variable
export LD_LIBRARY_PATH=${CUDA_HOME}/lib64
#Add CUDA_HOME to PATH
export PATH=${CUDA_HOME}/bin:${PATH}

OpenCV

https://gist.github.com/filitchp/5645d5eebfefe374218fa2cbf89189aa

Useful Tools:

• tmux
• htop
• gpustat