1. Machine Diagnosis
We create an artificial brain wherein the brain will be able to detect the symptoms. Then using past history, we will be able to predict a disease when the symptoms are known. Moreover, this can be extended (a little too ambitious) to a point where we will be able to predict times when the disease will occur in future even before the immediate symptoms are known.
This is used to gather data from the audience using facial recognition, gesture recognition, etc. Using that data, the speakers will be able to assess themselves better.
3. Self-driving car (Sponsored)
An intelligent agent is built for self-driving car simulation that efficiently detects road lanes, vehicles and traffic signs in real time using Canny edge, Linear SVM and Neural Nets. This is done using OpenCV alone and no expensive hardware except low cost GPU.
(GitHub private repo)
4. Virtual shopping using VR:
This is about online shopping. Here, user gets a 3D viewpoint of the product which he/she wants to purchase.
5. Linking UID to other parameters (mini-project)
In this, the users’ unique identification document (driver’s license in this case) was linked to all the documents to which the users were connected. This was useful in mapping all the user’s documents.