architsharma97/VisualQA

Stars
2
Language
Python
Created almost 9 years ago
Updated over 7 years ago

architsharma97/VisualQA

architsharma97

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

A model for Visual Question Answering. Extracts image features using the vgg16 architecture. Extracts language features using LSTM. Outputs from both are combined using a fully connected MLP, outputs of which belong to pre-chosen set of most probable answers.

dpo-rlaif

Jupyter Notebook

earl_benchmark

EARL: Environment for Autonomous Reinforcement Learning

VideoSummarization

Course Project for CS771: Machine Learning

medal

Code to reproduce results for MEDAL in PyTorch. Also contains code for running SAC and FBRL.

CV

Personal CV, based on Alessandro Plasmati's.

VisualDialog

The repository contains Theano implementations for Visual Dialog Models (http://visualdialog.org/). In particular, the Late Fusion and Memory Network models have been implemented.

mbsvme

Mixture of Bayesian SVM Experts

architsharma97.github.io

Personal Website based on "Goa" for Hugo

InstanceRetrieval

Assignment for CS698O: Visual Recognition

MNIST

Tests convergence of various estimators. In particular, PD and SF/REINFORCE are tested for gaussian and bernoulli latent variables and Synthetic gradients are tested in MNIST Classification (DNI).

ImageToolbox

A simple toolbox which does Red-Eye Reduction and Segmentation based Selective Blurring of Images. Has a simple Tkinter based GUI, which accepts images and performs the selected functions.

Privacy_SPMD

Privacy analsyis in DSRC: Data taken from Ann Arbor, Michigan. Conclusively shows that switching IDs randomly will not preserve user privacy is BSMs are transmitted at 10 Hz. Also shows lack of privacy if IDs are changed universally. Other protocols under test, where IDs are only transmitted when driving through an intersection.