Description I have implemented a Siamese Neural Network for text similarity. It takes two sentences as input and predicts the similarity between them. Here one input comes from the user and the other input is a entire dataset of documents in the data lake. It implements BERT embeddings for both the …

Objective Generate automatic Red Light Violation Tickets for Indian License Plate by detecting the License Numbers using Computer Vision. Description In this project, I was reading the CCTV feed(Images) of traffic Junctions, and performing the object detection task to identify the license plates …

Objective Build a NLP model which can extract out the important keywords (related to a particular domain) from documents.  Description In this project i have trained a custom Named Entity Recognition Model, which takes input of strings (max size 40) and predicts the Named Entities from the text. It …

Description I have written Selenium code in Python which can scrap any type of data from any website(dynamic as well as static). It uses Chrome Driver, and waits for given time till an object of a certain class is loaded completely. then it finds all relevant data related to a class name and appends …

Objective Predict the estimated time of arrival of a running bus at a bus stand with in-house intelligence. Description I have built a pyspark application which reads the current location and speed of a vehicle from an IVC (in-vehicle computer) and calculates the ETA using haversine distance …

Objective Provide a suggestion to select the optimal pricing tiers based on demand in a parking lot. Description II have trained a machine learning model using Prophet (Generalized Additive Model) which predicts the occupancy percent of a parking lot for next 3 hours. this forecast is used to …