Hi, I am Sat Patel, a Computer Engineering Graduate Student at Texas A&M.

This hybrid predictor is made up of the state of the art 64-byte TAGE predictor and the 2BC-gshare hybrid predictor which consists of a local and global predictor. 

Our novel neural network based chooser is implemented to decide which prediction to take by training a set of data on branch history and prediction accuracy. 

These training files are generated for each of the predictors by running traces through each of our branch predictors.

Supervisor: Prof. Daniel A. Jiménez, Computer Science and Engineering Department, Texas A&M University. 

Working on a hybrid DWT-DCT based frequency domain watermarking embedding operation. Currently testing different strategies (some DCT based, some DWT based, some based on both on DCT-DWT) on Python(using OpenCV, Numpy, Scipy). Testing the robustness of my system from three tests: 

1. Gaussian and Salt and Pepper Noise attack 

2. JPEG Compression 

3. Cropping and Scaling 

Once an effective model is finalized, I will work to implement the design on an FPGA and optimize it. 

Supervisor: Prof. Kishor Upla (Post-Doc), Electronics and Communications Engineering Department, SVNIT Surat.