We propose a novel spatio-temporal network to recognize micro-expressions from videos.

This paper is on detecting geographical landmarks in videos.

Tracking objects in motion can be used in many applications such as theft detection using CCTV cameras, and vehicle speed identification. One of the main issues with tracking objects is that whilst some methods are geared towards depthwise motion, the others are more effective in detecting movements across the captured frame. Another issue with most methods is their inability to detect slow moving objects. We propose a method which through experimentation has shown, could track motion of objects in all 3 dimensions as well as detect small, slow moving objects with minimal disturbance due to noise.

Aircraft stalls are when the aircraft goes into a uncountable state due to its angle of attack being too high and the power of the engines is not enough to fly the aircraft. The study produces a system to predict the probability of stall based on sensory inputs of the aircraft’s Speed, Altitude, Angle of attack and Thrust applied and have shown through testing that it may be used to correctly predict the probability of stall.

The classification of food according to their taste based on measurable features is important in ensuring an objective review of a food product. In this study, we use machine learning techniques such as Kth Nearest Neighbours, Support Vector Machines, Decision Trees and Random Forests to determine the quality of red wine based on 11 of its measurable features. We find that the RF models and KNN models perform better when compared to the rest.

The Mandelbrot set creates a visual representation of the function 𝒇(𝒙) = 𝒙𝟐 + 𝒄 by sampling the complex numbers and testing and shows a ever recursive pattern when zoomed. This document uses the round robin partitioning to parallelize the Mandelbrot algorithm with the use of Message Passing Interface (MPI). The performance of the parallelization is analysed using Amdahl’s law. The experimental results show significant speed ups with the use of parallel processing

In this paper, we propose to use a parallelized system for detection via the use of Message Passing Interface (MPI) in the C language for the communication between nodes as well as the use of OpenMP for the encryption of messages that are being passed. We aim to improve the efficiency in Inter Process Communication between nodes and as the results show, the use of MPI.