I am a Computer Engineering undergrad currently studying at SVNIT, Surat. I am a python developer with concepts familiarized in Django and handling the databases with Django . I have also tried out machine learning and computer vision and have made various projects in computer vision .
My core qualifications includes Leadership Skills , Patient learner , good communications and highly curious about ML and CV.
The following python files will be able to execute different filters . In snapchat , instagram , photobooth , etc there are many different filters which are based on facial recognition . The code is open source and as such , the code will execute a filter on your eyes . I have used opencv , dlib and numpy to create this filter .
View ProjectThe project comprises of three individual different tasks . Initially , it takes a picture into account and then a pixel-sort algorithm is applied to it after the picture is sorted , for every pixels its sound frequency is calculated and stored in an audio file . The last part of the project includes creating an audio-visualizer.
View ProjectEffects for different images including the negative effect , the 3D retro effect , technically know as anagylph and the sketch effect can be achieved through my code currently . These effects are similar to those of photo booth , snapchat filters and photoshop. All these effects will be saved in a separate folder.
View ProjectThe project was primirarly focused on developing a website and able to learn the server-side interaction using Django as the backend framework . The front-end of the website was accomplished using bootstrap only . Apart from this , for making the website responsive jQuery was used and AJAX request using jQuery was implemented to communicate on the sever-side .
View ProjectThe project used the object detection API to detect the object in a moving video . To track the detected object , the contour tracking algorithm was implemented with perspective transform and coordinates of the specific object detected were maintained in a CSV file as it were tracked in different frames . It was made to be a real-time application by using already trained weights .
View ProjectMulti class image classifier made using keras and tensorflow implemented on the Caltech 101 Dataset. This project focused on training the entire dataset first on a convolutional neural network and using weights of an already trained model(Inception-V3) .After reaching an accuracy of 90% + on both training as well as test dataset , the entire machine learning model was ready.
View Project