Professor Yao Wang, Department of Electrical & Computer Engineering, New York University

Professor Gaurav Sharma, Department of Computer Science & Engineering, IIT Kanpur

Submitted by: Vikulp Bansal

The project aims at using cube-map representation instead of equirectangular projection for rendering (reducing redundancy significantly) and using Dynamic Adaptive Streaming over HTTP (DASH) for developing streaming techniques.

Professor Ovidiu Daescu, Department of Computer Science and Engineering, University of Texas at Dallas

Professor Shashank K Mehta, Department of Computer Science and Engineering, IIT Kanpur

Submitted by: Sunil Kumar Pandey

We addressed the following problem :- Given an undirected , unweighted graph G with n vertices and 2n-2 edges, one of which is a double edge, find if G admits two edge-disjoint spanning trees and if yes, output them.

Professor Shaloo Rakheja, Department of Electrical and Computer Engineering, New York University

Submitted by: Saurav Thakur

The MVS 2.0.0 model is semi empirical model that incorporates the lowering of the virtual source (the peak of the energy band also known as top of barrier) charge due non-equilibrium transport. The project required some background knowledge that was mostly covered by course on nano HUB-U on nano-transistors and the rest was covered by other resources.

Professor Vibhav Gogate, Department of Computer Science, University of Texas at Dallas

Professor Piyush Rai, Department of Computer Science and Engineering, IIT Kanpur

Submitted by: Rushab Munot

The main aspect of the project was to study parameter tying by quantization used for regularization.

Professor Shivendra S. Panwar, Professor and Department Chair, Electrical and Computer Engineering & Director, CATT

Professor Adrish Banerjee,Associate Professor, Department of Electrical Engineering, IIT Kanpur

Submitted by: Rekhanshi Varma

Due to the widespread use of various wireless applications, wireless networks are becoming extremely important and hence require information security and stability during the transmission. Lately, physical layer security has turned out to be an emerging technique for remarkably improving the communication security of wireless networks.

Prof. Shivendra S. Panwar, Director, Center for Advanced Technology in Telecommunications,Professor, Department of Electrical and Computer Engineering

Amir Hosseini, PhD Candidate, NYU Polytechnic School of Engineering & NYU Wireless

Submitted by: Priyadarshi Anubhav

Most of the recently developed and widely used and created Virtual reality content is present in the form of Stereoscopic videos and 360° videos. These videos are usually consumed by the end users in their home where the all of the rendering and processing are done on high powered processing units such as dedicated GPUs for example in SteamVR.

Professor Nasir Memon, Department Head, Department of Computer Science and Engineering, New York University Tandon School of Engineering

Professor Piyush Rai, Department of Computer Science and Engineering, IIT Kanpur

Jay Koven, PhD candidate, Department of Computer Science and Engineering, New York University Tandon School of Engineering

Submitted by: Nishit Asnani

An email forensics investigator, when searching for evidence of fraud or illegal practices from a digital mail database, tends to rely on emails retrieved on a carefully crafted keyword search query. Then he/she goes through the list of results to find the required evidence, if it exists. This is a cumbersome process, when the database sizes are in the order of millions, and the relevant results are in the order of 10000s, which is usually the case.

Professor Nasir Memon, Department Head, Department of Computer Science and Engineering, New York University Tandon School of Engineering

Professor Sandeep Shukla, Poonam and Prabhu Goel Chair Professor, Department of Computer Science and Engineering, IIT Kanpur

Claude Fachkha, Postdoctoral Associate, Department of Computer Science and Engineering, New York University Abu Dhabi

Submitted By: Nikhil Vanjani

The Internet has become highly integrated into our current everyday lives. While cyberspace provides major benefits, our increasing reliance on it is producing new, significant vulnerabilities. As such, any security breach has the potential to result in debilitating effects on security, economy, public health, or safety. Furthermore, cyber attacks are dramatically increasing in size and number. Recent online threats demonstrated that organizations and governmental agencies could be subjected, nearly instantaneously and in full anonymity, to large-scale disrupting and orchestrated attacks with the potential to lead to severe security, privacy, and economic consequences (e.g., cyber-terrorism, DDoS, DRDoS, information theft, spam, fraud, child exploitation, etc.)

Professor Ovidiu Daescu, Department of Computer Science and Engineering, University of Texas at Dallas

Professor Shashank K Mehta, Department of Computer Science and Engineering, IIT Kanpur

Submitted by: Nayan Deshmukh

We addressed the following problem:- Given an undirected, unweighted graph G with n vertices and 2n-2 edges, one of which is a double edge. If G admits two edge-disjoint spanning trees and if yes output them

Professor Farshad Khorrami, Department of Electrical and Computer Engineering, New York University

Professor S. R. Sahoo, Department of Electrical Engineering, IIT Kanpur

Submitted by: Mayank Mittal

Autonomous Underwater Vehicles (AUVs), unlike other mobile robotics, are a relatively new field of research. They have helped humans conquer the dangerous waters, making tasks like data collection in marine environments, conduction of oceanic surveys, detection of mines possible.

Professor Nasir Memon, Department Head, Department of Computer Science and Engineering, New York University Tandon School of Engineering

Professor Sandeep Shukla, Department of Computer Science and Engineering, IIT Kanpur

Dr. Manoranjan Mohanty, Department of Computer Science and Engineering, New York University Tandon School of Engineering

Submitted by: Kunal Kapila

Photographs have been used to document reality and events since the invention of camera. Today, with the advent of photo and video editing softwares, it is easy to fake such data. Even though steps have been taken by having cameras sign the authenticity of images, we cannot ensure that the proof of authenticity of images is preserved even after basic transformations like cropping, rotation and compression. Existing methods try to solve this issue by having signatures that are remain invariant with respect to a few of the permissible transformations, but they are not sufficient. Photoproof, the main focus of this project aims to solve this problem by using cryptographic proofs.

Prof. Shivendra Panwar, Department of Electrical and Computer Engineering, New York University Polytechnic Institute

Amir Hosseini, Ph.D. Candidate, Department of Electrical and Computer Engineering, New York University Polytechnic Institute

Submitted by: Kartheek Rangineni

Video Streaming services occupy a significant share of mobile data-traffic because of their cost-effectiveness compared to Downloading services. Stream loading is a recently developed service which allow mobile users to enjoy download quality videos, while still being legally classified as a streaming service.

Professor Yao Wang, Professor, Department of Electrical Engineering, New York University Tandon School of Engineering

Yilin Song, PhD candidate, Department of Electrical Engineering, New York University Tandon School of Engineering

Submitted by: Kanishk Gandhi

The overall project is to enhance the experience for journalism by eliminating the redundant or repetitive information in one social event. Say there are many videos (or video shots) linking to one social event, we want to automatically cluster all videos have the similar semantic label or carry similar information.

Professor Nicholas Ruozzi, Department of Computer Science, UT Dallas

Professor Piyush Rai, Department of Computer Science and Engineering, IIT Kanpur

Submitted by: Ishika Soni

Structured SVMs are used to make structured prediction. Structured predictions are obtained by joint MAP estimation of output variables and hidden variables and discarding the hidden variables. Marginal Structured predictions involves marginalization of the hidden variables and then finding a MAP estimate.

Professor Yao Wang, Department of Electrical and Computer Engineering, New York University

Professor Gaurav Sharma, Department of Computer Science and Engineering, IIT Kanpur

Submitted by: Gaurav Verma

The project aims at developing rendering and adaptive streaming techniques of 360-degree videos by using cube-map representation instead of ER representation (and hence reducing the redundant information by a significant amount).

Professor Farshad Khorrami, Department of Electrical and Computer Engineering, New York University

Professor SR Sahoo, Department of Electrical Engineering, IIT Kanpur

Submitted by: Gaurav Sharma

In robotic mapping, simultaneous localization and mapping (SLAM) is the computational problem of constructing or updating a map of an unknown environment while simultaneously keeping track of an agent’s location within it. The project aims at exploring and implementing the SLAM techniques to map our environment using an Extended kalman filter.

Professor Vincent Ng, Department of Computer Science, UTD

Professor Purushottam Kar, Department of Computer Science, IIT Kanpur

Professor Piyush Rai, Department of Computer Science, IIT Kanpur

Submitted by: Divyat Mahajan

My research project dealt around a task in International Workshop on Semantic Evaluation 2016: Detecting Stance in Tweets. It involves automatic classification of tweets into two categories as Support and Against for a given target. A tweet can directly opinion about the main target, express opinion on some other entity apart from the target given or express no opinion at all.

Professor Vibhav Gogate, Department of Computer Science, University of Texas at Dallas

Professor Purushottam Kar, Department of Computer Science, IIT Kanpur

Submitted by: Amur Ghose

Sum Product Networks, henceforth SPNs, are a class of deep nets along the same lines as Bayesian networks but allow quick inference as well as computation of marginals. Most literature on SPNs focus on discrete SPNs and data sets, almost no cases go in depth into continuous cases. We aim to induce a SPN over continuous data sets using kernel estimation.

Professor Ovidiu Daescu, Department of Computer Science and Engineering, University of Texas at Dallas

Professor Shashank K Mehta, Department of Computer Science and Engineering, IIT Kanpur

Submitted by: Abhinav Garg

Given an undirected, unweighted graph G with n vertices and 2n-2 edges, one of which is a double edge. If G admits two edge-disjoint spanning trees and if yes, output them.

Professor Vibhav Gogate, Department of Computer Science and Engineering, University of Texas, Dallas

Professor Piyush Rai, Department of Computer Science and Engineering, IIT Kanpur

Submitted by: Abhinav Agarwal

Parameter tying is a new regularization alternative where parameters(weights) are tied after some learning is done using a similarity metric. There are additional expected benefits along with reduction in over fitting such as increased computational time and low memory requirements if this technique is applied with full effect.