Currently working on Data hub which creates a big data pipeline for transferring data instantly.

Big data Developer Intern

Developed a SocketTester and EchoServer for checking the Sentry logs which is similar to Samsung Knox. Tested and Automated the Android End to End Native Tunnel using NextGen Framework.

Software Developer Intern in Test

A fascinating issue in a digital forensic investigation is that given a digital video, would it be conceivable to recognize the camera model which was utilized to get the video. In the paper we take a simplified form of this issue by attempting to recognize recordings caught by a predetermined number of camera models. We propose various features which could be utilized by a classifier to distinguish the source camera of a video. We likewise give exploratory results and show sensible exactness in recognizing video for three distinctive camera models utilizing the proposed features. We have additionally attempted to improvise the time complexity for feature extraction using parallel processing.

Research Intern

Implemented a software defined radio using GNU companion software. Transmitted data from USRP to Telsob mote.

Intern

Worked at Sap PALO ALTO and Mobileiron in java environment. 

Publsihed an IEEE International paper in Machine Learning where I have used this software extensively.

Currently Working on hadoop at Sap palo alto.

A passionate programmer with strategic and focused approach towards finding solutions. Professional and creative. Love to explore new ideas.

Abhignana Mihir

 Implemented an 8 puzzle game with uniform cost search and A* algorithm with Misplaced tiles and Manhattan distance heuristics.
Implemented a nearest neighbor classifier with feature forward selection, backward elimination and own genetic algorithm for small data set of 24 features and large data set of 100 features. 

 
8 puzzle and nearest neighbor classifier

 Developed a Tweet search engine which also directs you to the twitter page and classifies every search result as positive, negative and neutral based on the text in the tweet. Used Hadoop and Lucene for indexing. Developed the web application using Rest API 

 
Tweet search engine with sentimental anaylsis

 Built a modern text editor that incorporates “intelligent” behaviors of modern-day text interfaces. Incorporated auto complete, flagging misspelled words and spelling auto-correction using java. 

 
Modern Text Editor

 Predicted hand written digits using logistic regression.
Designed a movie recommender system using collaborative filtering algorithm and anomaly detection.
Reduced the dimension of images using K-means compression.
Classified spam using Support vector machines. 
This  is a stanford course by Andrew NG

 
Machine learning projects

 The point in this project is to use a robot using which we can get photographs of a given thing from known edges. We mean to repeat and model these 2D photos to get the 3D sort of the article, which will be created and demonstrated in MATLAB. A 3D image reconstruction strategy from stereo pictures is displayed next that needs insignificant mediation from the client .The upside of this framework is that the stereo pictures don't have to be calibrated to acquire a remaking. Results for both the cam alignment and reproduction are displayed to confirm that it is conceivable to acquire a 3D model specifically from features of pictures 3D reproduction and demonstrating is utilized as a part of numerous fields like Virtual Reality, perceiving and controlling items and so on. 

Personal details

Abhignana M.

Summary

Technical skills

Work Experience

Education

Personal Projects