rishab-sharma
Full time Bayesian, part time Frequentist | BITSian
ISRO - Indian Space Research OrganisationNew Delhi
Pinned Repositories
chatbot_tensor
[https://rishabbot.herokuapp.com] I implement Tensorflows Sequence to Sequence model to train a chatbot on a reddit/AIML compatible open-source dataset. The bot holds a fun conversation and its just been trained for sometime .
cnn-hand-written-digit
This is the code for a model that recognizes handwritten digit images (MNIST). Developed using TensorFlow and the super simple Keras Library. Wrapped into a Webapp using Flask Micro Framework.
Cross-a-Crater
India has a prominent spot in the space research domain and stands tall along with the developed nations. It has added one more feather to its cap, when on September 26, 2016, India's space agency ISRO launched eight satellites from one rocket into two different orbits. ISRO has been leading the space research since its inception in 1969 and demonstrated its technical prowess by sending “MOM: Mission on Mars” on a minuscule budget. Taking inspiration from these missions, e-Yantra has developed the theme “Cross a Crater” for eYRC-2016. Consider the following scenario: e-Yantra has sent its rover for an expedition to the red planet Mars. The mission is to collect samples to determine if life sustaining factors exist in this planet. While returning to the Base Station from a different route the rover has encountered a huge crater that it needs to cross. The crater has two paths comprising of cavities along the way. These cavities need to be filled using appropriate boulders by taking a feed from the nearest satellite to make them traversable. The above scenario has been simplified and abstracted as an arena for this theme. The arena represents a crater and comprises of two partially traversable bridges, Bridge 1 and Bridge 2 with cavities at random positions leading to the Base Station. The rover takes the feed from a camera directly above it that guides it towards filling the cavities using conical structures and navigating the bridge. Navigating each of the bridges involve different challenges. The rover has to traverse using one of these bridges and reach the Base Station. The teams have to design, program and control an autonomous robot and use image processing techniques to complete the tasks. The team which traverses the bridge and reaches the base in the least possible time will be declared the Winner
emergecy_services
Web Portal For All Type of Emergency Services || They are the Police, Fire Brigade , Ambulance Service and Electricity. Emergency services are usually free. Also a special provision for our geek pals , we also provide a Internet emergency service which you can reach out for help on a attack like DDOS or SQLInjection.
geospatial_map_plotting
MAP PLOTTING USING GEOSPATIAL DATABASE is 2 way project in which a geospatial database can be plotted on a map depicting the various polygons which refer to the lease area of a company along with the point data depicting the wells and boreholes.
image_research
As observed machine learning, computer vision techniques and other computer science algorithms cannot compete the human level of intelligence in pattern recognition such as hand written digits and traffic signs. But here we have reviewed a biologically plausible deep neural network architecture which can make it possible using a fully parameterizable GPU implementation deep neural network independent of the pre-wired feature extractors designing, which are rather learned in a supervised way. In this method tiny fields of winner neurons gives sparsely connected neural layers which leads to huge network depth as found in human like species between retina and visual cortex. The winning neurons are trained on many columns of deep neurons to attain expertise on pre-processed inputs in many different ways after which their predictions are averaged. Also GPU used, enables the models to be trained faster than usual. Upon testing the proposed method over MNIST handwriting data it achieves a near-human performance. Upon considering traffic sign recognition, our architecture has an upper hand by a factor of two. We also tried to improve the state-of-theart on a huge amount of common image classification benchmarks.
nidaan
A webapp for disease recognition and symptoms analysis || Models : CNN , K-Means , Support Vector Classifier
ocr_on_android
An android app to recognise a hand Written Digit using a Flask , Tensorflow Backend , Model Designed using The coolest and simplest Keras
self_driven_future
My most ambitious project for simulating a self driving car with EEG. This is the code for my project where I used Udacity's self driving car simulator as a testbed for training an autonomous car.
snack_search
A machine Learning app for snack suggestion | The goal for this project is to build a system that allows you to identify and then recommend, recipes you're likely to enjoy.
rishab-sharma's Repositories
rishab-sharma/cnn-hand-written-digit
This is the code for a model that recognizes handwritten digit images (MNIST). Developed using TensorFlow and the super simple Keras Library. Wrapped into a Webapp using Flask Micro Framework.
rishab-sharma/emergecy_services
Web Portal For All Type of Emergency Services || They are the Police, Fire Brigade , Ambulance Service and Electricity. Emergency services are usually free. Also a special provision for our geek pals , we also provide a Internet emergency service which you can reach out for help on a attack like DDOS or SQLInjection.
rishab-sharma/image_research
As observed machine learning, computer vision techniques and other computer science algorithms cannot compete the human level of intelligence in pattern recognition such as hand written digits and traffic signs. But here we have reviewed a biologically plausible deep neural network architecture which can make it possible using a fully parameterizable GPU implementation deep neural network independent of the pre-wired feature extractors designing, which are rather learned in a supervised way. In this method tiny fields of winner neurons gives sparsely connected neural layers which leads to huge network depth as found in human like species between retina and visual cortex. The winning neurons are trained on many columns of deep neurons to attain expertise on pre-processed inputs in many different ways after which their predictions are averaged. Also GPU used, enables the models to be trained faster than usual. Upon testing the proposed method over MNIST handwriting data it achieves a near-human performance. Upon considering traffic sign recognition, our architecture has an upper hand by a factor of two. We also tried to improve the state-of-theart on a huge amount of common image classification benchmarks.
rishab-sharma/object_detection
A tensorflow model to detect objects in a webcam Video using OpenCV capable of localizing and identifying multiple objects in a single image
rishab-sharma/university_hub
University Hub is an App to connect the administration and the students through a simplified webapp . The Schema currently includes three different sections including the announcements section , the general discussion section and the semester forum .
rishab-sharma/codesutra_website
A website which I made for my startup Name: Codesutra , Its one of my best sample work for front-end , totally recommended by me if You want to see my work on front-end tech.
rishab-sharma/dressopidea
A machine Learning capstone app for dress searching and catalog browsing.
rishab-sharma/flask_practice_app
My sample work for a simple web-app made on Python micro-framework Flask(Backend Technology)
rishab-sharma/image_classifier
Image Classifier in TensorFlow || an extension of my research work
rishab-sharma/nodejs-express-practice-app
My First app on NodeJs Express (Backend Technology)
rishab-sharma/algorithms
Implementation of Algorithms in Python
rishab-sharma/lua_for_Vrep
Following are the non-threaded Child Scripts in Lua for V-Rep , V-REP is used for fast algorithm development, factory automation simulations, fast prototyping and verification, robotics related education, remote monitoring, safety double-checking, etc.
rishab-sharma/nmi_on_tensorflow
A Naval Mine Identifier Model Build on Pure Tensorflow Library , for testing tensorboard and its complications
rishab-sharma/serving
This is the code to Deploy a Tensorflow Model in Production using Tensorflow serving.
rishab-sharma/AIML_Chatbot
AIML: Artificial Intelligence Markup Language AIML (Artificial Intelligence Markup Language) is an XML-compliant language that's easy to learn, and makes it possible for you to begin customizing an Alicebot or creating one from scratch within minutes.
rishab-sharma/DAT210x-Microsoft-Project
My Data Science codes
rishab-sharma/deep_app
A deep analysis app
rishab-sharma/faceai
一款优秀的人脸、视频、文字:检测、识别的智能AI项目。
rishab-sharma/GANotebooks
wgan, wgan2(improved, gp), infogan, and dcgan implementation in lasagne, keras, pytorch
rishab-sharma/hint_repo
Hack in the north
rishab-sharma/image_repo
My Image Repository deployed on Heroku using Gunicorn with a flask Backend
rishab-sharma/julia
The Julia Language: A fresh approach to technical computing.
rishab-sharma/latex_for_one_page_two_col
Latex Code for document Preperation of Pdf with Two parallel Columns in one Page
rishab-sharma/local-development
Run Hasura locally on your computer
rishab-sharma/mit-deep-learning-book-pdf
MIT Deep Learning Book in PDF format (complete and parts) by Ian Goodfellow, Yoshua Bengio and Aaron Courville
rishab-sharma/nmt
TensorFlow Neural Machine Translation Tutorial
rishab-sharma/portfolio
HTML Code of my Portfolio - www.rishabsharma.co.nf
rishab-sharma/pratyeti
A navigation app for giving safest route , Smart SOS service and crime data visulaisation
rishab-sharma/recognition
A Webapp to detect face of Narendra Modi and Arvind Kejriwal using convolutional Neural Network
rishab-sharma/server
A server repo