Rajat Gupta's repositories
dog-project
Udacity Artificial Intelligence Nanodegree CNN Project. Created a CNN for breed classification of dogs by transfer learning using bottleneck features of four pre-trained models. Implemented the Haar cascade classifier for human face detection.
AIND-Isolation
Udacity AI Nanodegree's Project for a Game playing agent for Isolation. This project uses algorithms like minimax search, alpha beta pruning and iterative deepening to create a game playing agent for a zero sum board game like Isolation.
aind2-nlp-capstone
Udacity Artificial Intelligence Nanodegree Capstone Project. A Machine Translator for the translation of english to french. Trained four models Simple RNN, Embedded RNN, Bidirectional RNN and one with all three features with the later having maximum accuracy.
CarND-Advanced-Lane-Lines
Created a pipeline to identify lane lines using various techniques like perspective transform, color gradients, detecting fit for the lanes, detecting curvature and finally warping all to the original image.
CarND-Behavioral-Cloning-P3
Created a model to clone the driving behavior using convolutional neural networks. Trained, validated and tested the model using keras. The model outputs a steering angle to an autonomous vehicle.
CarND-Extended-Kalman-Filter-Project
Created an Extended kalman filter to estimate the state of a moving object of interest with noisy lidar and radar measurements. Calculated and obtained the RMSE values lower than the tolerance outlined in the project.
CarND-LaneLines-P1
Created a pipeline for finding lane line on the road. In this project I used python with OpenCV and used the hough transform to plot the hough lines in a region of interest.
CarND-Traffic-Sign-Classifier-Project
Created a pipeline to classify traffic sign images based on the German Traffic Sign dataset. Performed the exploration & visualization of the data; Designed, trained and tested the model architecture and Used the model to predict softmax probabilities of the new images.
AIND-Planning
Udacity AI Nanodegree's Project for implementing a Planning search. This project uses multiple search algorithms to solve a planning problem and then analyses the results based on certain metrics to find the best heuristic.
AIND-Recognizer
Udacity AI Nanodegree's Project for a Sign Language Recognizer. This project demonstrates the power of probabalistic models and employs hidden Markov models (HMM's) to analyze a series of measurements taken from videos of American Sign Language (ASL).
blog-template
Blog template for local hosting
boxreb14.github.io
Personal Blog using the Hyde theme built from Poole (the Jekyll butler)
bullet2test
Minimal test project using Bullet, cmake and git submodule
cartpoleplusplus
3d cartpole gym env using bullet physics trained from pixels with LRPG, DDPG & NAF
deeplearning-papernotes
Summaries and notes on Deep Learning research papers
Recommenders
Recommender Systems
Sunshine
Weather App build with the help from Udacity Android & Google #SunshineApp
tinyrenderer
A brief computer graphics / rendering course