Seasons Of Code

Computer vision based web app    • Senthil Vikram Vodapalli    • Dharshan   

WnCC - Seasons of Code

Seasons of Code is a programme launched by WnCC along the lines of the Google Summer of Code. It provides one with an opprtunity to learn and participate in a variety of interesting projects under the mentorship of the very best in our institute.

List of Running Projects

Computer vision based web app

Computer vision based web app

Object detection with live input & data analysis using tensorflow, opencv, yolo, faster-rcnn, python

No. of mentees: 4


In the early weeks, the mentee will be revising python programming, ML concepts focussing on computer vision and image processing. Next, the mentee will be implementing basic classification and recognition models and going through the state-of-the-art literature on detection models. Next, the mentee will work on implementing a transfer learning model of state-of-the-art object detection model and testing it on PC completely offline with live video input. The results from the detection model will be saved into a file and the mentee will work on building a web application and deploying the model to output results using flask with considerably good UI. Depending on the time available, further improvements can be made in the data analysis and UI part.


  • Image processing:
  • Basic web development:
  • Flask:

Tentative Project Timeline

Week Number Tasks to be Completed
Week 1 Brushing up python, ML and image processing.
Week 2 Implementing an image classification model (cat vs dogs)
Week 3 Implementing an object detection model and going through the state of the art literature in detection models.
Week 4 Working on transfer learning detection model of selected architecture in colab
Week 5 Making the model to work in offline and start working on web application
Week 6 Making a database out of the results and drawing inferences using python libraries
Week 7 Deploying the model results on the web application using flask
Week 8 Error handling, further improvements and back log cover-up week


Checkpoint Number Progress
1 Implement,validate and fine-tune a binary classification model from scratch for cat vs dog dataset from microsoft
2 Implement ,validate and fine-tune an object detection (on what classes/data) algorithm from scratch
3 Perform transfer learning on existing state of the art networks to fine tune them to our requirements
4 Build a web frame-work/app with user friendly UI for live, interactive user experience with the model of best performance.
5 Perform basic data analysis with the generated output of the model to determine useful information about the input data.