Given a list of latitude and longitude coordinates of airplane stops organize a path with minimum transportation cost.
Given a set of ‘n’ 2 dimensional points which are latitude and longitude values. There is a transportation company which has the task to pick all its customers with the help of ‘V’ vehicles with ‘C’ capacity each. The overall task was to minimize the overall transportation cost.
Big data analytics – building of inverted index using map reduce (core of hadoop).
Parallelizing clustering algorithms with map reduce for distributed systems.
Retrieve the data from social networking blogs. Develop the algorithms for analyzing the contents to detect unusual happenings like earthquake, landslide, tsunami, heavy rain. Identify the impact region and generate warning messages.
Predict whether a person is suitable for donating blood or not using few features.
Text mining on twitter dataset in JSON format. Using sentiment analysis to find out terms with very high significance in politics.
De-noising input question and effectively searching the correct answer of the noisy question.
Building a Span filter using SVM classifier and comparing it with Decision Trees (ID3).
Analysing loan data of Lending Club Association and prediction whether a person is eligible to apply for loan or not.
Analysing anomalies in bio-medical images using clustering technique.
In school there are different types of students, some studious, a few are silent, but mostly they are mischievous. Hence the principal decided to place CC cameras in the school. Design an algorithm which detects mischievous actions such as running, poking, shouting etc. [multiple class abnormality detection]
These days Smart TV’s have advanced features such as hand gesture recognition to change channel and increase and decrease the volume. Design an algorithm which detects gestures and assign specific actions such as increase or decrease volume to each gesture and detect and perform in real time. [real time action detection]
When working with tree bark and leaves images most relevant feature is texture. Study, compare and apply different approaches for texture analysis like GLCM, SFTA.
Using the Kalman filter to track and predict the motions of a randomly moving object in a video.
Using the particle filter to track and predict the motion of a randomly moving object in a video.
There is a famous game called as “The lemon in the Spoon Race” in which we need to hold the spoon straight in order to keep the lemon from falling off. Design an algorithm to detect the lemon in the video and track the lemon whether it stays on the spoon or not. [Object tracking]
Making the use of Hidden Markov Model for the recognition of hand movements and gestures.
The human hand consists of various parts such as wrist, fingers, and palm. For complete gesture recognition it is essential to determine the different aspects and track them.
Detection and classification of gestures that are a part of the Indian Sign Language. The project will make use of learning techniques for recognition of gestures from Images.
The project aims at detecting a hand in real time in a complex background.
The project aims at detecting and recognizing hand gestures using a single or combination of cameras.
The project aims at detecting and recognizing hand gestures using sensor based inputs like Microsoft kinect or data gloves.
Aims at recognizing non manual features of a sign language like facial expressions,body postures etc,which in itself can represent a sign or can modify the meaning of a sign
Detection of skin is tedious task. The project aims at developing a technique to detect the color of the skin making the use of various color models.
The value of color changes for different models under different camera and background settings. The project aims towards the development of color models that can identify precise shades of colors irrespective of the faults present in the Camera.
This is a part of the system which tries to allot stable color to objects when the illumination conditions are varying. As the distribution of the color intensities changes with the different illumination, different cameras etc. so it is required to reimburse these variations.
Background subtraction or foreground detection is applied where some objects are in motion and rest objects are stationary and we recognize the objects which are in the foreground. We have to compare performance of different approaches like Gaussian Mixture model, Otsu threshold etc. in term of false detection and noise produced by them. Try to build a new model for real time background subtraction.
Using an overhead camera to guide a robot to follow a specific path. The project requires an external camera attached to a computer/laptop as well as a two wheel differential drive robot that will communicate to laptop via serial communication .
Develop algorithms for analyzing multispectral satellite images for detecting and marking various regions like vegetation, roads, buildings, rivers etc. Further, implement the algorithm on multicore processors by extracting the parallelism for real-time applications.
Here we will study different morphological operation available and their application. Try to apply combinations of them to build approximate hand image from a distorted raw image. Try to make system intelligent as much as possible. For ex. System should tolerate scale, rotation variations and decide where an operation must be used so that best results are obtained.
To identify the different categories from multimedia content (Text, Image, audio, Video..etc) and then making the groups of the similar contents by using K-means clustering algorithm.
Before retrieving the information from the multimedia documents, categorize them according to their contents (clustering).
Making the clusters of various images according to their domains (based on the content of images) using k-means clustering algorithm that can be used to perform efficient image retrieval.
After making the clusters of various images, each cluster is annotated according to the domain related meta-tags so that the information is retrieved on the basis of these annotated tags.
To retrieve the information from the multimedia database (i.e. Text, Image, Audio and Video etc..).
It Deals with working on the Algorithms used to detect the shift in the glacier using Multispectral Sattelite Images Data. Like computing Glacier velocity etc.
It deals with the identification of Glacier Hazards such as Avalanches, glacial lake outburst flood and its impact on the vegetation.
Information Retrieval portal of best doctor in an area based on a particular health problem, using graph search.
Detect an employee of an organisation based on its uniques RFID code by a RFID radio receiver to get attendance, it can further be used to locate the location of the employee in the office premises. .
Online examination system, in which the level of questions increases/decreases based on the user perfromace. It also plots the learning curve of the user based on the test performance.
Developement of a mobile app that scans and decode a QR-Code to retrieve the encrypted information from it and process it in server.
Automatic time-table generator using activity selection.