PROJECTS
INFLUESENSE

Influesense is a portal for consumers to share and evaluate product and service reviews as well as rate them with respect to their peers. Influesense is a platform to formalize the product and service review segment so reviewers can be scored and consumers will easily be able to research before any purchase. Reviewers can share the same post on their respective social media profiles on the click of a button.

STAMPCAT

What is common between Warren Buffett and Anurag Srivastava you ask? Certainly not their respective net worth! Both are passionate about philately. Philately is the study, collection, research and appreciation of postal stamps and other related items. The first adhesive postage stamp was invented by Sir Rowland Hill in 1837, an act for which he was knighted.

Anurag combined his hobby for stamp collection and technology to create a digital platform that parses through an image of a stamp and identifies their origin, authenticity, value and heritage. The application is being developed using Python 3.7.3 with the support of Convolutional Neural Network algorithms, Flask API, TensorFlow and Keras predefined library.

Currently, the application is being trained for country specific data, although it is designed to be used internationally.


R API Interface

CNN algorithms for stamp identification Testing data from (ind1852)


Outcome:-

Dataset Example:-

Stamp dataset with appropriate name


Inside dataset have number of training data


After completing the dataset creation, we will populate the meta dataset which will translate and store the image features into corresponding numbers. This meta data will display the accuracy of the training data as well as test results from the training.



Data training with accuracy and validation result


After finishing with the training module, we will feed the system with sample data for evaluation. The parity of the sample images with that of the meta data will be displayed in decimals. So higher the parity, value will be closer to 1 or otherwise.


Testing with Australia 1982 stamp image


Test Results

We can see in the ‘Test Results’ that ratio of ‘australia1982’ is 0.998188 which is considerably greater than the rest of the datasets so the stamp is identified to be from Australia and produced in the year 1982.


ARTICLES

Taha Shaikh

Hi folks!!A few weeks back, I landed upon this tool called Figma merely by chance and believe that this is by far one of the best and wholesome design tools I have used so far and I’ve used a lot of them for one or other reasons. Currently I’m working with Figma for UI design, creating prototypes and collaborating with the development teams as well as Clients. I’m also able to version control and give limited view access to team members which is extremely handy as most of the projects are executed in an agile approach.

Being a cloud based browser application, I don’t have to bother about location specific accessibility or which device I’m accessing from, seems it runs equally well on my MacBook as well as Windows system. Earlier I used to work on Sketch so the learning curve to migrate from Sketch to Figma was pretty smooth and moreover, transferring files from Sketch to Figma is quite easy and hassle free. For design handoff I use Zeplin and HTML/CSS in Bootstrap Studio.