Portfolio

Checkout few of my Research

Illustration

Image Classification Using Bag Of Visual Words Model With FAST And FREAK - Best Paper Award

Neetika Singhal, Nishank Singhal , V.Kalaichelvi

This paper presents a novel technique of image classification using bag of visual words model. The entire process first involves feature detection of images using FAST, the choice made in order to speed up the process of detection. Then comes the stage of feature extraction for which FREAK, a binary feature descriptor is employed. K-means clustering is then applied in order to make the bag of visual words. Every image, expressed as a histogram of visual words is fed to a supervised learning model, SVM for training. SVM is then tested for classification of images into respective classes. The maximum accuracy obtained by the method proposed is 90.8%.

Read More

Application

Application of Convolutional Neural Network to Classify Sitting and Standing Postures - best paper award

Nishank Singhal, Srishti, and V. Kalaichelvi

—The paper aims at identifying whether a sitting or standing posture of a person is correct or incorrect using image processing and deep learning approach. The approach includes: (i) to check whether a person is present or not in the image read (ii) if present, then detect whether the person is sitting or standing (iii) in case of sitting, identify whether the posture is correct or incorrect (iv) in case of standing, identify whether the posture is correct or incorrect. In accomplishing the task, an overall accuracy of 91.3% is achieved. The method has been evaluated by testing it with a real time video feed thereby demonstrating the efficiency of the model and the wonderfulpower of Convolutional Neural Network (CNN).

Read More

Web Design

Comparing CNN and RNN for Prediction of Judgement in Video Interview Based on Facial Gestures

Nishank Singhal,Neetika Singhal , Srishti

This paper presents a novel technique of judging the performance of a candidate in a video interview. The candidate is judged as confident and attentive or unconfident and inattentive by taking the direction of face and eye into consideration. This corresponds to how many times is the candidate interacting actively, by making a firm eye contact with the interviewer. Image Processing techniques like Haar Cascade, Image filtering, Gamma Correction have been used for the detection of face and eye. Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) have been used for training and testing the images into right classes.

Read more

Application

Advances in Engineering and Information Technology A Level 4 Autonomy Self Driving Car Protocol for the UAE - Best Paper Award

Mr. Indraneel Patil, Mr. Juzar Gulamali, Mr. Nishank Singhal,Shivnarain Ravichandran, Dr. Abdul Razak

Roy Amara’s eponymous law famously states, “we tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.” In line with Dubai’s Autonomous Transportation Strategy, Dubai, Abu Dhabi and the rest of the United Arab Emirates is making proactive efforts to make 25% of the total transportation autonomous. The current self driving car (Autonomy Level 3) technology allows the Driver much comfort on the road through a combination of various AI and Machine Learning based algorithms on Park Assist, Auto Pilot and Cruise Control and advanced sensors for simultaneous localisation and mapping. As the world is gearing towards a Level 4 Autonomy Car, a lot of issues in the Cyber Security of the car, Intelligence and unpredictable driving scenarios on the road remain unaddressed. Our team through this research paper has tried to propose threenovel solutions namely Dynamic Region of Interest, Round-about Central Unit and Thermal Imaging Cameras for enhancement in the existing technology and then we implemented them to test their efficiency with our very own miniature prototype of a Level 4 Autonomy Self Driving Cars called Maverick. Based on our experience through this project and our knowledge of sensors and SLAM we have tried to extrapolate the technology used in the Maverick to the real world streets of the UAE

Read More