��ݮ��Ƶ

��

��

��

��

Many young and novice chefs often struggle to balance spices perfectly in their dishes, resulting in inconsistent flavors and wasted ingredients. Spice-it is an innovative, smart kitchen appliance designed to simplify cooking by accurately dispensing seasonings based on the user’s preferences and preset recipes. This eliminates the hassle of manually measuring and pouring multiple spices while keeping your hands clean.

PlantGuru is a smart home device that aims to solve the issue of plant care by reminding users when to water their plants. The system aggregates data from light, soil moisture, and temperature sensors and uses predictive modeling to determine a watering schedule using control theory. Users interact with the system through the PlantGuru mobile application, which reports their plants’ health. PlantGuru sets itself apart as a combination of a physical monitor and a smart cloud application in an all-in-one package.

Fleet management companies face significant annual losses due to vehicle downtime and accidents caused by poor driving habits. Imagine a solution that minimizes these losses by improving driver behaviour and monitoring vehicle health to prevent downtime. FleetVision is an all-in-one platform designed to optimize fleet performance. With machine learning powered real-time dangerous driving detection and advanced vehicle diagnostic analytics, FleetVision provides fleet managers with live insights into driver behaviour and vehicle health. This modern, data-driven solution enables proactive measures to minimize accidents and vehicle downtime.

Our project develops a programming library to simplify audio processing, focusing on enhancing human voices. Users can easily integrate operations like noise suppression and filtering. Using signal processing theory and Fast Fourier Transforms, our lightweight library prioritizes traditional techniques for fast and efficient performance. This versatile and accessible solution empowers users to tailor audio experiences to their needs, transforming how audio enhancements are applied across various platforms and devices.

Managing rental properties can be overwhelming for both landlords and tenants, from security concerns to rent collection and maintenance requests. Tenantra simplifies property management by bringing everything into one seamless platform. Secure keyless entry, easy rent payments, and a streamlined maintenance request system enhance safety, convenience, and communication. By eliminating the hassle of multiple disconnected systems, Tenantra ensures a smooth and stress-free renting experience—making home management easier and more secure for everyone involved.

Our project, HandSense, aims to design a hardware-based system for real-time hand gesture recognition using a glove with IMU sensors on fingers. The system will employ advanced signal processing to recognize both pre-defined gestures and gestures recorded by the user, offering an efficient alternative to camera-based methods, which could eliminate limitations like platform dependency, requiring multiple cameras, and privacy issues. Key challenges include ergonomic glove design, low latency, and a user-friendly interface. This approach eliminates visual input constraints and enhances integration with various platforms.

For native ASL speakers, over 99% of their interactions are with people unversed with the language, creating communication barriers and contributing to 40% unemployment among Deaf Canadians. This project aims to design a wristband that translates ASL gestures into spoken words. The Sign Echo uses IMU and electromyogram sensors to read hand movements, transmitting data to a phone via Bluetooth, which a machine learning algorithm translates.��

Deep neural networks like ConvNet, ChatGPT and DeepSeek consume enormous amounts of computation resources, yet with embarrassingly parallel and fixed workload. Models' architecture and weight remain unchanged for months after training. Our project exploits this predictability by capturing model architectures within a Field-Programmable-Gate-Array chip. Introducing Neural Networks on FPGA - preliminary work in a "one-model, one-chip” design where an entire fully connected layer and��convolution blocks are kept within the chip. Achieving the goal to eliminate most intermediate memory transfers for extraordinary computation power and efficiency on inferencing.

SwipeCommand addresses the common issue of managing multiple TV remotes by eliminating the need for physical remotes. SwipeCommand processes these gestures into commands for TVs and streaming devices by using a camera to detect user gestures. The system uses machine learning to translate gestures into infrared signals and other various transmission methods, achieving a user-friendly experience with a response time under 0.5 seconds. This innovative solution combines computer vision, signal processing, and embedded systems to create a novel universal remote alternative.

NavWay aims to create an end-to-end platform for automating indoor navigation mapping with a smart robot. The lack of available indoor navigation solutions hinders businesses and visitors in indoor spaces like malls and offices. Using Structure from Motion (SfM), NavWay generates detailed 3D maps and provides accurate navigation for users through an augmented reality mobile app. Unlike other alternatives like Google Maps, NavWay offers an automated, interactive, easy-to-use solution for precise indoor navigation.

Many people would like to enjoy keeping houseplants but lack the experience to care for a plant properly. PlantPulse is a smart home plant monitoring system designed for house plant enthusiasts seeking personalized care. Using IoT sensors, it tracks soil moisture, sunlight, temperature, and nutrients. Data is analyzed in the cloud against optimal conditions for common houseplants, delivering tailored care instructions and timely alerts via a user-friendly mobile app. Unlike generic automatic watering systems, PlantPulse offers explicit, customized guidance to ensure your indoor garden thrives.

Want to master your basketball jump shot? Then the Buzzer Beater arm sleeve is the perfect tool for you. Powered by a network of motion sensors, advanced AI vision models, and seamless integration with the Buzzer Beater app, jump into real-time training visualizations, personalized feedback, and expert tips from your favourite superstars. Buzzer Beater takes your game to the next level, and our innovative technology has the potential to revolutionize not only basketball training but also a wide range of sports, AR/VR applications, and more.

MLoV is an acceleration framework for deploying and adjusting deep learning ONNX models on hardware to optimize performance metrics.

To reduce the cost for inference, this library offers flexible hardware mappings to target a heterogeneous collection of field-programmable gate arrays and SoC development boards at the user’s availability. This enables a streamlined, plug-and-play experience for developers while abstracting hardware knowledge. By leveraging an open-source compiler and a distributed network, this project provides an efficient and affordable alternative for achieving performance per watt over host CPUs.

The ArithmaTrix is a children's puzzle toy designed to teach math and improve motor skills. With declining test scores, this 2x2 rotating cube, resembling a Rubik's Cube with screens, can enhance classroom learning. It consists of eight segments, each featuring three e-ink screens displaying numbers or arithmetic operators (+, -, x, /). Each cube face presents an equation tailored to the child's grade level. Once a child solves a face by rotating the cube, it grays out the face until the entire puzzle is completed.

Current insurance models fail to reflect individual driving behavior. Existing tracking solutions lack accuracy, preventing fair assessments of safe drivers. TrackSafe is a plug-and-play device using computer vision and sensors to monitor eye focus, speed, acceleration, and proximity. It provides real-time feedback via a mobile app, generating an algorithmic safety score to lower insurance rates. Our goal is to promote safer driving and fairer premiums by offering a comprehensive analysis to help drivers improve their habits.

The growing demand for enhanced home security has led to the proliferation of smart security cameras. However, these systems often lack intelligence and flood users with unnecessary notifications from any motion. By harnessing sensor fusion of camera and IR sensors, along with continuous adaptive learning, SmartSentry distinguishes between familiar and unfamiliar people, delivering reliable alerts. Furthermore, SmartSentry ensures privacy by exclusively training the system on only authorized users' data. This intelligent system minimizes unnecessary notifications and enhances home security with accuracy, privacy, and efficiency.

EdgeNode aims to simplify the deployment of IoT devices��and enhance their performance in the field. It seamlessly interacts with local, offline devices, sensors or microcontrollers and offloads ingested data to the broader network on demand. Its fully FPGA-based implementation ensures faster and more efficient operation.

PestPaladin is a smart pest monitoring system for homes, businesses, and agriculture. Using computer vision and machine learning, it provides real-time detection, identification, and tracking of pests through smart cameras. Compatible with existing home security systems, webcams, and custom IoT setups, PestPaladin offers home and commercial users flexibility, delivering info through mobile and web apps with real-time notifications, visualizations, and analytics.

AutoGreen is a smart plant care system designed to automate the maintenance of a single plant. It uses sensors, a microcontroller, and components like a water pump, electric heater, and vent to monitor and regulate temperature, soil moisture, and watering. A mobile app provides data visualization and control, making it ideal for users seeking minimal effort in plant care or those growing plants in challenging climates. AutoGreen ensures optimal growth conditions, combining technology and convenience for plant enthusiasts.

Increasing residential demand, including electric vehicle charging, strains the existing distribution system, causing transformers to age prematurely and fail if not upgraded. TxVision addresses this issue with a two-part solution: a meter for transformers that transmits current consumption and service outages data and a machine-learning model that forecasts demand trends. This helps utility companies predict future growth in specific residential areas, enabling targeted upgrades, efficient resource allocation, improved grid reliability, and better integration of renewable energy sources.

Every year, almost 3 million sports-related concussions occur, with 50% of those concussions going unreported or undetected. As well, research has shown that even small hits to the head over time can lead to brain dysfunction in athletes. Our project aims to help detect and track trauma to the head from sports injuries that may result in concussions. A detection system that tracks a player's trauma sustained over time will allow early preventative measures to ensure player health and safety.

A smart waste-sorting bin that uses AI and computer vision to automatically classify and sort trash into recycling, compost, or garbage. Users simply place an item inside and press a button. The bin captures an image, sends it to an AI model for identification, and directs the waste to the correct compartment. With automated mechanisms, it streamlines waste management, reduces sorting errors, and promotes sustainability.

The firmware and software of commercial heart-rate sensors are proprietary, which restricts data integrity and the ability to reproduce results. Nevertheless, open-source hardware platforms like Arduino and Raspberry Pi offer system integration. Unfortunately, there is a gap in the market for validated and easy-to-use heart-rate analysis algorithms, especially for low-cost PPG sensors that provide noisy data. Poorly documented or undocumented research implementations that require more than average technical skills are too common. This project intends to develop an open-source algorithm that is accurate, robust, and applicable for real-time heartbeat detection.

Have you ever listened to a song and wanted to play it with your favourite instrument, but can’t find the sheet music anywhere? You no longer need to transcribe music by ear with Cadenza tediously. Cadenza is a new web service that turns audio recordings into sheet music in seconds. Cadenza utilizes AI to split songs into vocals. Then Cadenza uses signal processing algorithms to identify notes, and sheet music is constructed within minutes!

Proper footwork is critical for success in sports, but it isn’t easy to practice properly without instructors present. MobileCap is a wearable motion tracking system that displays a 2D projection of your movement. The required algorithms are computed right on your mobile device. The visualization on your mobile device allows you to pause, slow down, or speed up the playback for analysis. MobileCap is a low-cost and easy-to-use inertial motion capture solution, enabling you to capture your motion exactly where you practice.

Contemporary neural network signal processing relies on quantizing a signal in space and time and then performing the inference. This quantization of large volumes of input data before performing feature extraction adds latency and is computationally inefficient. The Analog Neural Signal Processing Unit is a Printed Circuit Board with discrete digital and analog circuits that can perform the inference in the analog domain with continuous signals, resulting in space, time, and power savings resulting from quantizing on the feature level instead.

Struggling with dandruff, hair loss, or scalp irritation? pHPerfect is a smart accessory that detects your scalp’s pH levels and sends real-time results to a mobile app. The pHPerfect app analyses pH measurements over time and delivers personalized hair care recommendations, including tailored product suggestions. The pHPerfect app connects users through a supportive community. ��Easy to use, and cost-effective, pHPerfect empowers you to take control of your scalp health —no expensive treatments needed. Healthy hair starts with a balanced scalp!

Approximately 300,000 Canadians and 5 million Americans rely on wheelchairs daily. Of these individuals, 85% suffer from soft tissue problems, most of which are pressure ulcers. Thus, there is a clear need for a system for those with mobility issues to monitor and decrease the prevalence of pressure sores. Under Pressure is an auto-inflating cushion that intelligently redistributes and relieves load from high stress areas, reducing the chance of developing sores and allowing people to monitor historical data for better health management.

The Standup! is a compact and highly practical device that helps combat the adverse effects of prolonged sitting by reminding users to stand at regular intervals. Studies show that excessive sitting is linked to a 34% increased risk of early death for adults sitting 10 hours per day. With intuitive gesture controls, adjusting settings like timer duration is effortless. It features a long battery life and can be worn on the wrist or carried in a pocket. Subtle vibrations provide effective reminders, promoting movement, reducing sedentary habits, and improving overall well-being.

Ternary Accelerator is engineered to boost inference speeds in large language models. Designed for the same purpose as prior acceleration systems, it replaces conventional multiplications with simpler additions and subtractions, streamlining the core arithmetic operations. This innovative approach reduces computational overhead and energy consumption by leveraging FPGA hardware. The result is a more efficient process for generating high-quality language outputs, significantly enhancing the performance of applications such as conversational AI and automated translation. Ternary Accelerator offers a practical, cost-effective solution for modern LLM-driven tasks.

Experience a voice-activated document scanner enhanced by an extensible toolkit for swift, intelligent captures! Developers can integrate hardware sensors, streamline voice-enabled interfaces, or harness RAG pipelines with multi-modal LLMs, embeddings, and vector databases. Customize text extraction rules and compliance protocols or rebuild modules via APIs. It is ideal for education, legal, and assistive tech, it deploys seamlessly across cloud, edge, or hybrid systems, offering��turnkey efficiency and open customization for context-aware document automation.

Many public locations such as museum exhibits and information kiosks utilize audio to convey information to users, but the sound can disturb those nearby. BeamBox aims to create a system capable of broadcasting a targeted sound beam in specific directions to reduce disturbances and limit audio interference. BeamBox utilizes a phased array of ultrasonic speakers emitting an upconverted audio signal where directionality is controlled, and steering is performed to allow the concentration of sound beams to their desired locations.

Air pollution is often overlooked, despite being the fourth leading risk factor for early death. This project aims to develop a portable, battery-operated device to monitor air quality, featuring a display and a long battery life. Additional features including data analytics, alerts, and power profiles are accessible through a mobile app. Equipped with various sensors, the device provides real-time data and alerts for high pollutant levels. It combines ease of use with highly accurate sensors, creating a valuable and practical tool for monitoring air quality.

Traditional wireless communication systems can be prone to interference in high EMI environments, such as industrial settings with high-power machinery or medical settings with sensitive test equipment.��

Our project seeks to provide an alternative to laying fiber optics by using a much more resilient part of the spectrum - infrared. Using an FPGA-based modem, we adapt traditional RF modulation schemes, but using specialized amplifiers and an LED and photodiode instead of antennas, benefitting from EMI resilience and line-of-sight security.

The Safety Lane is an advanced robotics system that uses traffic cone robots to manage construction zone barriers remotely. The system leverages Bluetooth Low Energy to connect robots and the operator. In addition, Ultra-Wideband technology is used to ensure precise cone spacing. Operators can position and retrieve traffic cones using a mobile application, thereby reducing the need for construction workers to be near traffic and improving overall site management and safety.��

IllumiNote is a virtual piano teacher that provides an interactive, intuitive learning experience. Users select a song in the companion app, and the corresponding piano key LEDs light up in sequence, guiding them through each note. IllumiNote detects key presses using a microphone and provides feedback on accuracy. By eliminating the need to look between a screen and a piano, IllumiNote makes learning more seamless and accessible to all skill levels.

CarGuard mitigates car theft with advanced sensors and real-time notifications. Thousands of vehicles are stolen annually due to weak security. CarGuard’s Arduino-connected sensors detect suspicious activity, instantly notifying users via Bluetooth Low Energy (BLE) and requiring two-factor authentication (2FA). Unlike easily replicated fob RF signals, BLE connections are more secure. Authorized phones reconnect the relay, restoring power. This solution combines wireless communication and microcontroller programming for robust, responsive anti-theft protection.

This project aims to develop a smart pickleball paddle that provides players with real-time performance data, addressing the rapid growth and evolving needs of the sport. The paddle will integrate advanced sensors to capture crucial metrics about a player's game. Data will be transmitted to a mobile app for analysis, offering insights on shot placement, power, and technique. This innovative solution provides instant, accessible feedback to players, allowing them to understand their play style and improve their skills without the need for expensive coaching.

Studies have shown that the number of reported musculoskeletal pains associated with working from home has increased dramatically since COVID, reaching 3.2 million in 2023. PostureSense is a system that detects the user’s posture through their webcam and delivers real-time visual feedback. Using a combination of machine learning-driven image detection, data processing, and a physical LED array, our solution offers a non-invasive and cost-effective way to improve your posture while you work.

Invenstory is revolutionizing warehouse management with a data-first, vehicle-centric approach to inventory tracking. Over 75% of companies lose sales due to poor inventory control—our low-cost, scalable solution ensures that never happens again. Each vehicle in the warehouse acts as an independent node, transmitting real-time metadata for package tracking from arrival to departure. Invenstory uses WiFi fingerprinting for precise indoor tracking, making warehouse management accessible to all businesses. By transforming real-time data into actionable insights, Invenstory optimizes warehouse operations and drives efficiency, accuracy, and accountability.

Currently, the GPU market is dominated by two leading semiconductor companies - NVIDIA and AMD. The lack of competitors has reduced researchers, educators, and hobbyists' access to these products. Introducing OpenCore, built on AMD’s RDNA 2.0 ISA, the GPU is implemented on a FPGA with a custom PCB for handling power and communication. Our group aims to challenge the status quo by providing an open-source and free alternative to the existing graphics and AI processing chips.

SmartGrow is an autonomous greenhouse system that aims to mitigate premature plant death by optimizing growth conditions. It improves indoor plant survival rates and promotes healthier plant growth in a practical and user-friendly manner. Using image processing and convolutional neural networks, it identifies plants and optimizes lighting and soil moisture. Key features include automatic irrigation, ambient light control, and real-time monitoring on an LCD screen. SmartGrow integrates knowledge from image processing, computational intelligence, power electronics, and circuit theory for a user-friendly solution.

LifeGuard empowers seniors to live independently with a smartwatch designed for their unique needs. It features a simplified UI, real-time health tracking, and predictive alerts using machine learning to detect potential emergencies. Emergency alerts can notify caregivers and bystanders, ensuring rapid assistance. A code on the watch grants medical personnel immediate access to health records. The mobile app enables remote monitoring for caregivers. By prioritizing accessibility, proactive health monitoring, and emergency response, LifeGuard enhances safety and peace of mind for seniors and their families.

CargoBuddy is an autonomous robot that allows the elderly suffering from mobility issues to transport large items like groceries within their home. It uses sensors and pathfinding algorithms for navigation with an adaptive map and a spacious cart for larger items. Additionally, it contains obstacle avoidance to prevent potential harm and move items safely. Users can set delivery spots and monitor the robot’s location with an easy-to-use application. CargoBuddy is a practical solution for the elderly to improve their independence and quality of life.

InfrAlert aims to enhance driver safety in low visibility scenarios through an object detection system that utilizes thermal imaging to alert drivers of surrounding wildlife or pedestrians. Users will be able to see information on the type of objects they pass by to be better prepared for potential hazards in the future. InfrAlert reduces the cost of installing thermal detection systems directly from the manufacturer and is adaptable for any vehicle.

Parking in busy areas is often frustrating and time-consuming. ParkEasy is a smart parking system designed to make the process quicker and more efficient. It detects open spaces and provides up-to-date parking availability using cameras and real-time processing. This reduces the time spent searching for a spot, eases traffic congestion, and improves overall convenience. Whether in a crowded city or a packed parking lot, ParkEasy helps drivers find a space faster, saving time, fuel, and frustration.

VR/AR headsets lack the experience of touch sensations, causing a less immersive experience. Current haptic solutions are bulky and costly. Our project aims to design affordable and comfortable force feedback gloves that enhance the VR/AR experience by providing tactile sensations, offering precise hand and finger tracking at a lower cost and smaller form factor.

6ixSense VR bridges critical gaps in patient care during transfers by enabling doctors to be virtually present in real time, no matter their location. Using a gantry-mounted 360° camera with mechanical tracking, our system mirrors user’s VR headset movements for unified navigation. This empowers healthcare professionals to guide paramedics during emergencies, enhancing decision-making and saving lives. Beyond healthcare, 6ixSense VR transforms industries like remote training, entertainment, and disaster relief, which in turn pushes VR’s boundaries to create real-world impact.

Easy Breezy is an affordable, energy-efficient smart vent system designed to optimize home airflow while reducing heating costs. With home automation projected to reach $715.6 billion by 2032, inefficient ventilation remains a major issue, leading to inconsistencies in temperature and higher monthly bills. Our intelligent solution regulates airflow, improving comfort while minimizing energy costs. Unlike other solutions, Easy Breezy offers seamless and cost-effective integration for homeowners while enhancing efficiency, sustainability, and overall living conditions.

Student safety on field trips has always been a concern for both teachers and parents. The TripTracker system consists of portable real-time tracking devices and a smartphone app that monitors the distance between teachers and their students, alerting teachers if a student wanders too far away. Utilizing a combination of wireless positioning technologies, namely Ultra-wideband and GPS, a reliable connection is maintained when transferring between outdoor and indoor settings, helping teachers to react quickly and ensure a child’s safety.

To enhance occupational safety and efficiency, we introduce Nautilus—a remote-controlled firefighting vehicle designed to address these challenges, it is featured with an automatic fire detection system that combines deep learning, image processing, and real-time video transmission, integrated with self-designed PCBs and commercial antenna modules. Nautilus is dedicated to facilitating efficient information collection and survivor rescue within hazardous zones, providing a low-risk, cost-effective solution for firefighting teams.��

Cell phones are vital in serving as lifelines in emergencies for navigation or calling 911, especially in remote areas with limited access to power. However, battery life limits continuous access to these services. The FreeCharger aims to use Piezoelectric material to harness energy from daily phone movements to charge a battery on the back of a phone, providing emergency power when outlets aren't available. This design leverages natural movement, avoiding the need for external power sources.

As smart home technology advances, keyless locks are becoming standard in hotels and apartments, but frequent battery replacements remain challenging. Our project, Open Sesame, is a battery-less NFC-enabled smart lock that harvests power from a smartphone’s RF signal. Using an integrated antenna, power circuit, and microcontroller, it authenticates users and unlocks without requiring maintenance. Designed for efficiency and security, it eliminates battery waste while enhancing user experience. This innovation aims to make smart access systems more sustainable and reliable.

The bird-watching community is rapidly growing, with more than 50 million North Americans feeding wild birds. However, enthusiasts may not get to observe their favourite birds. Other animals often steal bird food, enthusiasts aren’t notified when birds are perched on the feeder, and bird seed may run out anytime. This project aims to design a low-power bird feeder that sends video clips to a user’s phone. By notifying users of bird visits, providing clear video and automatically dispensing bird seed, bird enthusiasts will never have to worry about missing their favourite birds. Using a low-power motion sensor, a MOS switch activates a camera when an animal approaches. Machine learning enables animal identification, wireless communication provides real-time notifications, and low-power circuit design ensures optimal battery life. With this new design, users experience several advantages such as bird food not being stolen by other animals and real-time notifications. With its solar-powered, energy-efficient design, WingWatcher enhances the bird-watching experience while ensuring minimal maintenance.

OptiDrone aims to demonstrate the applicability of neural networks to drone flight and allows users to simulate and test drone motion in a virtual environment, allowing for efficient visualization and testing while preserving user safety and drone integrity. Using a custom joystick controller and pre-programmed test plans, our team has trained a model to predict the drone’s motion and designed a control system to correct flight paths automatically. The data can be used for environmental planning, evidence collection, and civil engineering photogrammetry.

HearWise is an advanced hearing aid that enhances speech clarity while reducing background noise. It operates independently without external connectivity, using a scaled-down machine-learning model for accurate speech recognition and effective noise cancellation. By dynamically adjusting audio processing, HearWise optimizes conversational clarity, making it a reliable solution for individuals with hearing impairments. Its intelligent noise suppression technology delivers a seamless listening experience in various environments.

ScreenTickle is an attachable, touch-enabled frame that transforms non-touch screen devices into touch-responsive devices using infrared (IR) sensors and photodiodes. This frame��offers an affordable and seamless upgrade, allowing users to enjoy touch functionality without investing in new hardware. The system detects interruptions in the IR light beams and wirelessly connects to the computer. ScreenTickle is user-friendly, requires no professional installation, and can be easily attached or removed.

Conventional alarm clocks often struggle to get the job done: nearly 70% of people use the snooze button, averaging 22 extra minutes of sleep. The Runaway Alarm Clock is designed to tackle this problem by forcing the user to catch the device when the alarm sounds physically. Equipped with a microcontroller, LCD display, loudspeaker, motors, wheels, and distance sensors, it moves around the floor unpredictably while avoiding obstacles. This physical activity helps users wake up faster and effectively overcome sleep inertia.

Quantum computing presents a new avenue for solving problems impossible to solve with classical computers. The latest experiments require sub-nanosecond control over lasers, cameras, and other implements. Experimental setups are ever-changing, so we have constructed an open-source, one-click solution that generates firmware for every new configuration. This lets researchers update and adapt their experiments by giving them control over their hardware and eliminating their reliance on third parties.��

LightLink is a low-cost visible light communication (VLC) technology that integrates into existing lighting infrastructure to provide high-speed data transmission. By leveraging the unlicensed visible spectrum, Lightlink addresses the spectrum limitations of traditional wireless technologies due to the global surge in data traffic. Utilizing advanced signal processing and low-noise circuits using commercial RGB LEDs, photodiodes, and microcontrollers, LightLink establishes visible wavelength channels to improve bandwidth efficiency and reliability, addressing the growing demand for efficient wireless communication.

Navigating legal documents for housing issues is daunting for students. LegalEase, a virtual consultant, provides swift, authoritative guidance on housing rights, referencing court cases and the Residential Tenancies Act. This service saves students time and legal expenses for straightforward queries by offering transparent, verifiable citations. LegalEase aims to enhance awareness, prevent exploitation, and empower students to make informed decisions, thereby improving the rental landscape. By offering accessible and reliable legal guidance, LegalEase fosters a better-informed student community.

Visual impairments can make text hard to read, with some impairments causing text to be blurry. ReadSpex provides a wearable solution to help users read and understand text via text-to-speech to help navigate everyday life. This portable device also features a companion app to help users conveniently remember and recall the text they encounter.

Piracy costs the live-streaming industry billions, and current detection methods can’t keep up. Our live streaming piracy detection system leverages Machine Learning to analyze hardware behaviour, distinguishing between legitimate viewers and those illegally redistributing content. Unlike traditional approaches that rely on expensive manual monitoring, our system enables real-time, scalable enforcement. By helping content delivery networks protect their broadcasts more effectively, we provide a smarter way to safeguard digital content.

Today, many services require providing them with ultra-sensitive information directly (think copies of your passport or SIN and government ID cards). TIVS designs a decentralized system allowing users to verify their identity without exposing their personally identifiable information. Through trustless cryptography paradigms and a custom NFC-enabled terminal, TIVS proves the validity of official ID documents while ensuring users retain complete control of their sensitive details. We aim to fundamentally reimagine how identity verification can work in a privacy-first digital world.

Transforming books into immersive multimedia experiences has traditionally been expensive and out of reach for most authors; Book-to-Movie (B2M) reimagines this process by combining book scanning hardware with AI-driven technology, making it fast, affordable, and accessible.

B2M leverages AI-powered text-to-speech, image generation, and text segmentation to bring stories to life, assigning distinct voices to characters and generating visuals that match the narrative. This seamless integration allows authors to effortlessly convert their books into high-quality movies, expanding their reach and transforming how stories are told.