Vehicle Analytics: Proof-of-Concept Implementation at a Community Mall
By Mustavi Sarker, Chawalit Khotwong & Sorasit Arsirapongphisit
Vehicle Analytics a powerful traffic monitoring tool developed for the shopping mall StadiumOne. Vehicle Analytics was designed to leverage the power of deep learning technologies like Convolutional Neural Networks (CNN) to detect, classify and track vehicles passing by the road and calculate the road’s traffic density, along with the additional feature of keeping track of cars at parking lots. All data collected is uploaded to the cloud in real-time, then showed on a beautiful dashboard in real-time.
Various Cases from Market Research Perspectives
By Harpreet Singh Arora
Topic modeling, an unsupervised machine learning technique that was implemented in order to discover intent behind the collections of data. The purpose of using Topic Modeling was to aid Market Researchers in order to reduce their burden from a mandatory process of exploring the data to discover insights. Topic modeling was utilized on offline surveys to understand the context of responses by the people with additional visualization techniques applied to help Market Researchers understand survey responses in a simple, fast, and efficient manner.
ISL Retail Vision
Towards Building a Smart Retail Solution
By Avanish Shrestha & Utsaw Siwakoti & Wanthita Peaugpoolpol
A video analytics solution designed to help retailers to increase sales by learning more about their customers.
Piano Score Sheet
Pitch Alphabets Generator using Optical Music Recognition
By Menh Keo & Aung Khant Oo
Optical Music Recognition (OMR) has become a trend of study with increased demand of digital sheet music. In this showcase, we explore techniques and algorithms to implement optical music recognition. Our main purpose is to design an OMR system by using machine learning and computer vision to recognize the piano music notation and display the pitch alphabets back on the input sheet music to help the piano players who started in learning piano.
Tic Tac Toe
Simple AI Strategies for Competitive Tic Tac Toe
By Thanitsak Leaungsupornpong
A tic tac toe game with an AI that can play at any board size which to be exact is a size value that starts from 3 by 3 until a size that an integer class can hold. The same rule of tic tic toe applies to every size of the board which fills up all the horizontal, vertical, or diagonal lines. This game is program in C# with my personal rule based technique for AI.
Connect-4 AI Using Minimax & Alpha-Beta Pruning
By Aman Upadhya
Connect four game against an AI that is using minimax and alpha-beta pruning ( decision rule algorithm used Artificial intelligence and game theory) along with some heuristic that helps the AI to make its next move by evaluating the board and assigning score for each move .
How to Apply Heuristics to Create a Smarter Search Algorithm
By Dao Trung Kien
This is a fun project which solves the Sudoku, a famous logic based, combinational number placement puzzle, by using an AI. The AI includes the implementation of searching algorithms and heuristics to improve the solving performance and to think in the most “natural thinking” way. The project can read the sudoku from the screen and solve it automatically.
An Evolutionary Approach to a Tetris-playing AI
By Sochivoath Chiv
A fun little project, combining two spectrums of science into one AI bot. By applying Charles Darwin’s Theory of Evolution, the Tetris-playing AI continuously evolves and improves through a genetic algorithm in the game of tetris without needing any external intervention or supervision.
ISL Brand Intelligence
ISL Brand Intelligence, Text Analysis & Classification for Brand Management
By Dr.Kwankamol Nongpong & Shirish Kayastha
An Intelligent Content Analytics Solution.
Deep Reinforcement Learning
Deep Reinforcement Learning for Playing Atari Games with Large Number of States
By Jan Duldhardt