AI Projects #6 for Social Good Showcase

inzva AI Projects #6 for Social Good Showcase will take place on Saturday, May 8, from 13.00 to 16.15 via Zoom.

AI Projects is a 3-month research and project group where AI enthusiasts from different backgrounds come together to work on projects of their choosing with the GPU we provide for our participants.

AI is rapidly becoming a prominent part of the social impact and responsibility world, especially within the context of the United Nation’s Sustainable Development Goals. Projects developed under the title of “AI for Social Good” attract attention from both researchers and institutions dealing with societal problems.

As a non-profit that aims to utilize tech knowledge for social good, inzva has decided to hold the sixth batch of our biannual AI Projects with a theme and invited teams that want to tackle problems that will bring out the good in AI research with a variety of areas of focus including medical solutions, Turkish natural language processing applications and understanding a neighborhood’s demographic makeup by using street view images.

You can see the last previous showcase videos here.

Application period is closed.

Duration: 5 teams; 15 minutes presentation & 5 minutes Q&A

#6 PROJECTS

1- Turkish Offensive Language Detection

Abdurrahman Dilmac, Ekrem Bal, Talha Çolakoğlu, Uras Mutlu

Modern NLP tools with deep learning can be used to build a strong model that detects offensive language. Our aim is to reproduce the OffensEval 2020 benchmark on Turkish offensive language detection tasks, and then to improve the model with different techniques such as text normalization and data augmentation.

We plan to finalize the project as an open-source API so that interested people can contribute and those who need can use it in their applications or websites.

Hate speech and offensive language detection.png

2- Medical Diagnosis Decision Support System for Parkinson's Disease

Can Bulguoğlu, Duygu Ay, Handenur Çalışkan, Orhan Ağaoğlu

For generations, assessment of speech abnormalities in neurodegenerative disorders such as Parkinson's disease (PD) has been limited to perceptual testing or user-controlled laboratory analysis based on very small human vocalization samples. However, using speech signals, detection of Parkinson's disease is now possible with the help of machine learning and deep learning methods. In this project, our aim is to detect Parkinson's disease by processing speech signals and reach the state-of-the-art level. Furthermore, we want to investigate whether our model can achieve a generalized performance for the voice signals from different languages.

Medical diagnosis decision support system for Parkinson disease.png

3- Optimal Data Augmentation for Biomedical Image Segmentation

Billur Selin Zaza, Enes Sadi Uysal, Mehmet Yiğit Özgenç, Onur Boyar, Şafak Bilici

In this study, we aim to find the optimal data augmentation volume and optimal data augmentation strategies in the biomedical image segmentation problem. Data augmentation plays a crucial role in performance of most of the image classification on the image segmentation tasks. It is unavoidable when there is a shortage in the training samples. In the context of biomedical image segmentation, U-Net family architectures are widely used. In order to increase the performance of the model, most of the applications rely on the data augmentation. In this study, we are investigating the data augmentation techniques and their effects on the model performance. In addition, finding the optimal amount of the samples generated using these techniques is another problem we addressed in our study.

4- Object Detection with Street View Images for Demographic Analysis

Alara Hergün, Başak Ekinci, Efehan Danışman, Sefa Kurtipek

A large number of resources are allocated for data collection and estimating these structures to estimate demographic structures of the neighborhoods. As it has been done in other countries, we would like to use object detection using Google Street View Images of Istanbul, in order to predict socioeconomic development levels of each neighborhood. We will be determining objects in street view images such as cars, trees, and miscellaneous street items in order to predict the socio-economic development level of the neighborhood. Mahallem Istanbul project’s results are selected as target values to predict by using street view images.

Object detection with street view images for demographic analysis.png

5- Building Knowledge Graph-Based on Turkish Wikipedia

Alaeddin Selçuk Gürel, Ateş Bilgin, Mustafa Barış Çamlı, Okan Çiftçi

Knowledge graphs are basically a way to store knowledge on a graph form, using entities, relationships, and attributes. In this project, we will construct a knowledge graph using a not-so-common language data source which is the Turkish Wikipedia Dump. We will use the data in full texts and information boxes to get entities, relationships, and attributes. We will implement Language Models are Open Knowledge Graphs paper to extract information from full texts. We will also scrape information boxes with BeautifulSoap and improve the graph.

Building Knowledge Graph based on Turkish Wikipedia.png

Schedule for the day

13.00-13.15: Welcome

13.15-13.35: Project 1

13.40-14.00: Project 2

14.00-14.15: Coffee Break / Breakout Rooms

14.15-14.35: Project 3

14.40-15.00: Project 4

15.00-15.15: Coffee Break / Breakout Rooms

15.15-15.35: Project 5

15.40-15.55: Coffee Break / Breakout Rooms

15.55-16.15: Closing Remarks

All participants have to abide by our Code of Conduct and Letter of Consent

A BEV Foundation project inzva is a non-profit hacker community organizing study and project groups as well as camps in the fields of AI and Algorithm; and gathering CS students, academics and professionals in Turkey.

Subscribe to our newsletter here and stay tuned and follow us on our social media accounts to have the most recent news about our upcoming events and programs!

All personal data you shared in this form will solely be used to determine the participants of the event and to inform the venue provider, and will be deleted as soon as we carry out our legal obligations. Therefore, by sending this form, you accept having your personal data processed and transferred to third parties that are providing services for the event.