Aritra Banerjee

Aritra Banerjee, Michael Clear, Hitesh Tewari

Cryptocurrencies have received a lot of research attention in recent years following the release of the first cryptocurrency Bitcoin. With the rise in cryptocurrency transactions, the need for smart contracts has also increased. Smart contracts, in a nutshell, are digitally executed contracts wherein some parties execute a common goal. The main problem with most of the current smart contracts is that there is no privacy for a party's input to the contract from either the blockchain or the other parties. Our research builds on the Hawk project that provides transaction privacy along with support for smart contracts. However, Hawk relies on a special trusted party known as a manager, which must be trusted not to leak each party's input to the smart contract. In this paper, we present a practical private smart contract protocol that replaces the manager with an MPC protocol such that the function to be executed by the MPC protocol is relatively lightweight, involving little overhead added to the smart contract function, and uses practical sigma protocols and homomorphic commitments to prove to the blockchain that the sum of the incoming balances to the smart contract matches the sum of the outgoing balances.

Aritra Banerjee

Measuring the similarity between patches in images is a fundamental building block in various tasks. Naturally, the patch-size has a major impact on the matching quality, and on the consequent application performance. We try to use large patches instead of relatively small patches so that each patch contains more information. We use different feature extraction mechanisms to extract the features of each individual image patches which forms a feature matrix and find out the nearest neighbor patches in the image. The nearest patches are calculated using two different nearest neighbor algorithms in this paper for a query patch for a given image and the results have been demonstrated in this paper.

Aritra Banerjee

In this thesis the aim is to work on optimizing the modern machine learning models for personalized forecasting of Alzheimer Disease (AD) Progression from clinical trial data. The data comes from the TADPOLE challenge, which is one of the largest publicly available datasets for AD research (ADNI dataset). The goal of the project is to develop machine learning models that can be used to perform personalized forecasts of the participants cognitive changes (e.g., ADAS-Cog13 scores) over the time period of 6,12, 18 and 24 months in the future and the change in Clinical Status (CS) i.e., whether a person will convert to AD within 2 years or not. This is important for informing current clinical trials and better design of future clinical trials for AD. We will work with personalized Gaussian processes as machine learning models to predict ADAS-Cog13 score and Cox model along with a classifier to predict the conversion in a patient within 2 years.This project is done with the collaboration with researchers from the MIT MediaLab.

Aritra Banerjee, Michael Clear, Hitesh Tewari

Zero-knowledge proofs have always provided a clear solution when it comes to conveying information from a prover to a verifier or vice versa without revealing essential information about the process. Advancements in zero-knowledge have helped develop proofs which are succinct and provide non-interactive arguments of knowledge along with maintaining the zero-knowledge criteria. zk-SNARKs (Zero knowledge Succinct Non-Interactive Argument of Knowledge) are one such method that outshines itself when it comes to advancement of zero-knowledge proofs. The underlying principle of the Zcash algorithm is such that it delivers a full-fledged ledger-based digital currency with strong privacy guarantees and the root of ensuring privacy lies fully on the construction of a proper zk-SNARK. In this paper we elaborate and construct a concrete zk-SNARK proof from scratch and explain its role in the Zcash algorithm.

Aritra Banerjee, Dimitrios Kollias

The main idea of this ISO is to use StarGAN (A type of GAN model) to perform training and testing on an emotion dataset resulting in a emotion recognition which can be generated by the valence arousal score of the 7 basic expressions. We have created an entirely new dataset consisting of 4K videos. This dataset consists of all the basic 7 types of emotions: Happy, Sad, Angry, Surprised, Fear, Disgust, Neutral. We have performed face detection and alignment followed by annotating basic valence arousal values to the frames/images in the dataset depending on the emotions manually. Then the existing StarGAN model is trained on our created dataset after which some manual subjects were chosen to test the efficiency of the trained StarGAN model.

Aritra Banerjee, Shrey Choudhary

In this paper we have modified the existing page ranking mechanism as an advanced Page Rank Algorithm based on Semantics Inlinks Outlinks and Google Analytics. We have used Semantics page ranking to rank pages according to the word searched and match it with the metadata of the website and provide a value of rank according to the highest priority.We have also used Google analytics to store the number of hits of a website in a particular variable and add the required percentage amount to the ranking procedure.The proposed algorithm is used to find more relevant information according to users query.So this concept is very useful to display most valuable pages on the top of the result list on the basis of user browsing behaviour which reduce the search space to a large scale.