Beirut AI Community Recap: January 2021
There’s always something exciting going on over at Beirut AI and this January was packed with valuable experiences entirely online! Lockdown can sometimes be fun!
The spotlight was on Beirut AI Bootcamp which not only provided students and professionals with a solid foundation in artificial intelligence but also gave them a chance to get all that knowledge into outstanding projects during the AGRI hackathon!
Gathering with all the Enthusiasts 🚀
We were all anticipating Beirut AI Bootcamp and getting to meet more than 80 enthusiasts in the gathering preceding the Bootcamp. Many came from different backgrounds but with a similar goal: becoming AI heroes!
The participants were introduced to the Bootcamp slack channel to further connect with each other, share knowledge, and get to find a team and an idea for the upcoming hackathon everyone was looking forward to. It was an opportunity to be part of a community that aims to learn and grow despite the challenges. With all the resources, new connections, and a Bootcamp Boarding Kit, everyone was ready to dig deep into AI starting from Day 1!
Day 1: Data Science and Machine Learning 🚀
The first day was the beginning of a promising week! Participants got introduced to AI and its various types, main components, and applications made possible by its different fields.
The Data science life cycle was then explained in detail to build up for the discussion of machine learning and its framework. After all the theory, it was time to jump on Google Colaboratory and get hands-on experience through building a model to predict medical treatment cost based on various data!
Day 2: Deep Learning 🚀
As the Bootcamp started to pick up the pace, we were getting more excited to dig deeper into AI! It was time to learn more about deep learning and neural networks. This came with an introduction to TensorFlow and Keras followed by plenty of interesting concepts and examples.
The day was not just about AI, but the participants had a fun activity during the break: 2 Truths, 1 Lie! It was a unique way for everyone to get to meet one another in breakout rooms by sharing some fun facts about themselves while creating a friendly atmosphere. It was truly a boost that everyone needed to jump deep into the hands-on workshop and start building, training, and evaluating deep learning models!
Day 3: Computer Vision 🚀
“The science that helps machines see!”
Participants were introduced to computer vision and image processing as well as one of the most important computer vision libraries, OpenCV, with a practical overview of its capabilities. It was time for some “image math”!
Combining knowledge from the previous discussion about deep learning, we started exploring convolutional neural networks and Keras library’s role in computer vision.
Day 4: Natural Language Processing 🚀
How can machines understand what we say? What are some challenges faced? Where is NLP used? How can we integrate deep learning in NLP?
With these intriguing questions and many more, it was time to explore this interesting AI domain. From data preparation to Bag-of-words to recurrent neural networks to many other interesting concepts, there was a lot to learn and apply!
One fun way we applied our knowledge of NLP was by extracting people’s opinions from IMDb movie reviews!
Our enthusiasts were also up for a mini-game on Kahoot during the break. With intriguing trivia questions for everyone to play around with, there were those that really enjoyed this mini-game and got in the top 3 places! The excitement for the AGRI-Hackathon got us in a fast-paced mode during the game!
Day 5: Time Series Analysis 🚀
“Time for some Dynamic Data!”
As the Bootcamp came to an end, participants got introduced to time series analysis, the types of problems involved in it, TS datasets, TS models, and much more! Wrapping up with an interesting case study was a great way to apply all that was learned.
And finally, graduation day is always a day to be grateful for!
AGRI-Hackathon: Inspirational Projects 🚀
“AGRI-Hackathon kick-starts. The countdown begins!”
As the Bootcamp came to its last day, brilliant minds came together to build AI solutions for a green Lebanon!
The element of excitement that comes from a challenge was always apparent with even the smallest examples, like solving the Rubik’s cube, to bigger challenges, like building the project prototypes for the hackathon.
Hackathon participants benefited from Berytech’s Business Model Canvas workshop as well as technical and business mentoring sessions. This was to aid them with their pitches and help enhance their projects while getting valuable insights and leveling up their skills! They also had the chance to apply to Agrytech’s accelerator program and take their ideas to the next level!
As the closing ceremony of the hackathon was around the corner, 15 teams out of the total 45 teams got selected to compete in the live pitches and showcase their AI solutions to problems in the Lebanese agriculture sector. This was their opportunity to win one of the top 3 cash prizes and many valuable rewards such as the opportunity to gain free access to Beirut AI Weekend workshops happening in March — stay tuned for that!
There was something special that amazed us in every team, yet three teams impressed our judges a lot! Watch the closing ceremony here to find out why!
First Place | Green Vision 🚀
Most of our vines in Lebanon are prone to many diseases, and this is a very common phenomenon. To aid local farmers as well as find a way around harmful and costly pesticides, Green Vision team worked on an AI software that relies on Computer Vision to detect these diseases when simply given the needed image. The model they built is currently capable of distinguishing between tomatoes and different types of grapes, identifying if the crops are healthy or not, and suggesting ways to take care of the crops. Their software can be an aid in other fields such as robotics.
Second Place | Techponics 🚀
Do you want to grow your own crops? Thought about hydroponics?
Team Techponics created an app that gives users guidelines on setting their own hydroponics setup, access to a community, access to planting kits market, and the ability to monitor plants’ needs. Relying on Computer Vision and Deep Learning, their app is capable of identifying whether plants are healthy or not while providing other diagnostic data to the user. Their model was trained on scanning roots of known plant types to give care recommendations and suggestions.
Third Place | Bovifit 🚀
Team members: Elie Bou Faysal, Omar Awad, Robert Matta, Yasmina Skaff
Bovifit team targeted cattle farmers with their machine learning-based software with the aim of gathering daily vital data logged by farmers. Their model is capable of noticing minor data variations to predict heat phases, suitable insemination periods, and milk yield of cows. This freemium model software will help cut down the costs on the farmers’ side.
More from the Community 🚀
Are you interested in being part of our vibrant community?
We have something very special planned for you! Follow us on our social media accounts, subscribe to our newsletter, and stay tuned to make the best out of your experience with Beirut AI.