CSE Fall Departmental Demo Day
We’re thrilled to invite you to the fourth annual Comp. Sci. & Eng. Fall Demo Day. Student groups from several CSE capstone classes will be presenting the culmination of 3-months of effort, hard work, (metaphorical) blood, sweat (well… caffeine really), and tears (see above).
Where:
Davis Hall, 1st Floor Atrium
When:
Friday Dec. 7; 12 Noon - 3 PM
Schedule
11:00 -
Judges and Presenters arrive
11:30 -
Pre-event mingling opportunity for Judges and Presenters
12:00 -
Event opens to the public; Lunch is served; Judging starts
2:00 -
Judging ends
3:00 -
Awards presented
Acknowledgements
Thanks to everyone (Participants, Sponsors and Guests) for a hugely successful demo day! Check out the photos Courtesy of Ken Smith.
Sponsors
Awards
Entrepreneurs Award
-
UB CarpoolTeja Maddipatla, Ajaysai Potluri
Student's Choice
-
Caching for ARAlexander Beloglazov, Sandeep Kumar, Venkata Krishnan Anantha Raman, Vikram Singh
First Place
-
Trading AlgorithmsKaushik Panneerselvam Veerappan Saravanan
Second Place
-
Word By WordChristian Coffey, Daniel Goings, Daniel Sarlo, Doug Achkar
Third Place
-
HD Map Creation for AcademiaArun Krishnamurthy, Srivatsa Manjunath Hegde
Presented Projects
This year’s participating classes and projects include:
Software Engineering (CSE 442)
- Word by Word
-
Word by Word is our speed reading app that is designed to help people increase the rate at which they can read text and documents. Text can be input by the user manually or through image-to-text recognition.
- QR Attendance
-
Take student attendance with randomly generated QR code by prof/student sign information and attendance is stored in back end fire base data server
- DopAt - Classroom attendance and quiz website
-
- Mark Armstrong
- Eithne Amos
- Abass Ibrahim
- Connor Reynolds
- Chris Saless
The next 🦄
Masters Project Development (CSE 611)
- UB UNIFIER
-
- Vijaya Harshavardhan Palla
- Kalyani Sunil Chikte
A web application that brings all UB related events, club information under one hood keeping UB students up to date. It includes an extension that allows UB students to add events found online to their calendar.
- Social Media Broadcasting
-
- Vikram Singh
- Mahalakshmi Maddu
- 199 Admin Site
-
- Charanya Sudharsanan
- Christopher Chan
- Home Buyers and Sellers
-
- Komlan Aziagba
- Buffalo Museum - Tours
-
- Parush Garg
- Mariyah Husain
- Anjali Sujatha Nair
Buffalo Museum Tour App allows the museum visitors to take self guided tours from a variety of tour options available. The application provides navigation on the museum's map and information on various exhibits. It also provides an option to turn the tour into a quest and earn points for it. The application is targeted to all age groups with a friendly UI.
- Museum Events App
-
This app targets the general public of Buffalo to send information about different events happening at Buffalo Museum. The app sends push notifications to the mobile users about the latest and upcoming events. Users can also look for more information and purchase tickets for the events inside the app. Basically it is a platform for the Museum Staff to reach out to public in a subtle and convenient way. This app is developed with Native Android platform. So it is only available on Google Play store. Cheers :D
- Gamify - Buffalo Museum of Science
-
Gamify is a framework to allow users to sign up to earn "badges/points" for playing games. The framework has two games namely Quiz and Scavenger Hunt. These earned "badges/points" can be seen by others in the website that would allow users to compete with each other, and perhaps have a leader board to show who has the most "badges/points". This will encourage youth visitors to explore new areas of the museum to amass "badges/points".
- Happiness Experiment
-
- Sourabh Bhagat
- Pradeep Aitha
Happiness is the scoreboard for life. Happiness now, happiness at the end of the day, happiness with life. There is no other word in the English language that is as rich, deep, meaningful in regard to how we view our lives than happiness. To be happy our life needs to “feel right.” Our life needs to be close to in balance physically, mentally, emotionally and spiritually. We are the experts of our own happiness. You control your happiness. You can build a happiness routine that sets you up for success in life and “A life well lived.” Leading research has determined that 50% of our happiness is genetic, 10% circumstantial and 40% is “our intentional activity.” In other words our beliefs, mindset and routines. We hope that this program will inspire and help you take ownership and action to enable your happiness. Rather than read hundreds of self-help books and listen to podcast after podcast we have boiled the advice down to a simple (minutes a day) program that you can start, today. Right Now. World happiness, healthiness and greater good starts with the individual – yes, you!!! Once you have built your personal happiness then you can start sharing it with others creating a virtual cycle of happiness. The happiness experiment is a program that has been designed to bring more happiness to your life through playful experimentation. We are going to load up your day with happiness. Help you build a happiness routine. Inspire you to push your happiness boundaries.
- UB CarPool
-
Ub CarPool is a simple web app which connect two or people going to same destination for sharing their ride. Users can post, search and confirm rides using this app.
- Trading Algorithm: Looking for Predictive Value in Twitter Data
-
The goal of this project is to predict the likely tendency for SPY ETF, using the previous day’s Twitter data. This project aims at implementing a prototype which could be used to predict the trend of SPY ETF using tweets which could be used to design profitable trading algorithms."
- Secured Publish-Subscribe System
-
A system level implementation like MQTT for a secured Publish-Subscribe system and deploying it on cloud, making use of multiple instances.
- Anomaly Detection using Deep Learning
-
- Vikrant
The project involves understanding the LSTM deep log learning method. Then we have to implement the deep learning model using Python, Keras/Tensorflow, Pytorch etc ML frameworks. The project requirements include server infrastructure hosted in UB.
The project envisages:- Develop a solution to apply DeepLog to our system log data (from Lonestar supercomputer)
- Implement the solution and report results to identify performance anomalies from system log data
- Human Action Detection and Recognition
-
- Roshni Murali
- Aayush Kumar
The next 🦄
- Motion flow/depth estimation using synthetic video data
-
- Sili Liu
- Local IOT Cloud
-
- Kshitij Goel
- Moni Kumari
The next 🦄
- Object detection and pose estimation in temporal point clouds
-
- Bhavik Gala
- Charushi Nandwani
The project addresses the problem of real-time object detection in point clouds (LIDAR sensor data) in the context of autonomous driving. We have implemented an end to end object detector which detects the location and heading of the other cars with respect to self, this is crucial for the safety of the passengers, vehicle, and other vehicles on the street. The model can be easily adapted to detect pedestrians, cyclist, and other vehicular types given a dataset. The project is based on PIXOR detector and works at the speed > 10 frames per second.
- Distracted Driver Detection Service
-
This is a service that is designed to detect distracted driving. It is capable of accepting images as input, process it, and send back a response indicating whether the driver is distracted or not. The dataset for the model was obtained from Kaggle.
- Consensus Algo
-
- Raman Keswani
- AI based Health Assistant
-
- Anant Gupta
- Scheduling Framework
-
This product is aimed at developing a scheduling framework. The stakeholders are the professors who would like to schedule the TA’s recitation class timings for their courses. The primary focus is to implement a framework as a web application with the goal of scheduling the timings of TA’s recitation classes. With this product, the rules and constraints can be configured into the system and an interactive user interface will simplify the process of manual scheduling by dynamically guiding the user to set up the schedule without any conflicts. This product has been chosen to solve the long-standing problem of identifying and resolving the conflicts during manual scheduling."
- Credible Fandom
-
A new social network for sports fans where they can not only interact with one another, but also place predictions for sporting events to show off their knowledge of sports and develop their credibility score.
- Choreographic Lineage
-
- Shailesh Adhikari
- Shivam Sahu
The next 🦄
- Web of trust with social graph
-
The project aims to try to find the ‘best’ strategy to find possible answers for Universal fact based boolean questions based on user responses in a trust weighted directed graph.
The accuracy of the responses compared to the actual answer under different topologies of user responses will be used to define ‘best’ metric.
- TamedPy - Sandboxed Python Execution
-
A framework to run untrusted python workloads on modern infrastructure, in a safe sandboxed way. From interactive notebooks for Data Science to code execution for programming contests or hiring, run someone else's python code on your infrastructure without risking harm to your host environment.
- Decentralized IoT
-
- Paritosh Mandar Walvekar
- Aditya Kohli
An orchestration manager for a decentralized IoT application. Instead of IoT devices relying on cloud for control and actions, they should be able to communicate with each other and decide on a possible actions given a set of inputs/events.
Advanced Computer Systems (CSE 622)
- ORB-SLAM
-
- Ankesh Bhoi
- Swati Nair
- Rohan Pathak
Split the vision pipeline between the edge device (mobile phone) and the cloud and comparing the performance of the split system with the current one running completely on edge device
- Gesture Recognition on Secure World
-
Using TrustZone enabled board, we demonstrate a strongly secured gesture recognition that cannot be forgeable.
- Caching Service for AR Applications
-
Build a caching service that delivers a better user experience by pre-fetching the Map portions, the service thinking of a very thin client which doesnt require more compute which has connectivity to a server. Localization and Caching will be done on the thin client by exploiting the resources on the server.
- Software Verification
-
- Anirban Chatterjee
The next 🦄
Connected and Autonomous Vehicles: Applications and Systems (CSE 740)
- Modeling wireless channel
-
- Foad Hajiaghajani
Modeling wireless channel and effect of obstacles in network simulator 3 (NS-3) based on WAVE 802.11p wireless communication
- Object detection using convolutional neural network
-
- Yuhao Du
The next 🦄
- Simulating a LiDAR PointCloud in Unity
-
- Srinivas Ravi
- Raj Vinodchandra Soni
The next 🦄
- Security for Autonomous Vehicles
-
- Venugopal Shah
The next 🦄
- Localization
-
- Aakash Gupta
The next 🦄
UB Hacking
Computational Linguistics (CSE/LIN 467/567/667)
- Drug-Drug Interaction Extraction from Drug Labels
-
- Chinmay Swami
- Kumkum Kaushal
- Gursimrat Singh
- Detecting Kansai-ben
-
- Anirudh Manjunath
- Vivek Raghavan
- Ensemble Classifier for SQuAD 2.0
-
- Bowen Chu
- Lingbo Hu
- Yizhan Wu
- American Movie Recognition Project
-
- Sheng Liu
- Annotation of Multi-word Expressions in the French Treebank (conference submission)
-
- Wenqi Li
- Assigning proficiency levels to text of a learners corpus (journal submission)
-
- Niharika Raut
- Movie Recommendation
-
- Alex Kim
- DongKyu Lee
- Panpan Zhang
- Ce Zheng
- Sarcasm Detection
-
- Maggie Liu
- Jasmeet Kaur Chawla
- Controllable and Personalizable Chatbot
-
- Erin Pacquetet
- Deep Reinforcement Learning for Neural Machine Translation
-
- Tyler Hanavan
- Sentence classification
-
- Kurt Staufenberg