Are you curious about A.I. research? Would you like to build your own cutting-edge A.I. technology? Would you like to get published? If you answered ‘yes’ to any of these questions, then our advanced AI Research program may be right for you!

WHO:

Our advanced AI Research Program is offered to students with a strong interest in publishing their own AI research article. Our target audience are gifted students, undergraduates, and graduates. The prerequisites are pre-calculus and Python coding. Our curious and motivated students have 70+ publications and contributed lectures with IEEE, SIAM, AMS, and more and have been named finalists for both Regeneron STS and Regeneron ISEF.

WHAT:

Students will use AI technology to build the world’s fastest computer simulations. Past simulations have included tsunamis, coastal erosion, room acoustics, sonar, gene expression, wind tunnels, bridge design, fracture mechanics, carbon nanotubes, and more. Students can also use their simulations to inform an AI decision-making process.

WHEN:

We have availability in the evenings and weekends. We are on PST.

WHERE:

All lectures and tutoring will take place over Zoom.

WHY:

The best way to learn AI/ML is by completing your own fun and interesting project! Additionally, publishing your research will allow you to improve your resume and network with other scientists.

HOW:

Your instructor Derek will demonstrate graduate level techniques using only Google Colab and Microsoft Excel! No other material is necessary!

TUITION:

Undergraduate projects: $2400 for 16 hours of Zoom tutoring.

Professional projects: $4000 for 16 hours of Zoom tutoring.

OPTIONAL ITEMS:

These are services for your convenience:

• Bibliographic summaries: A great way to learn is to read research articles. To help you, we will provide short summaries of eight of the latest papers in your field.
• Cloud computing: We will run and monitor your scripts on our cloud account.
• Latex formatting: We will format your research article using LaTeX editing software.
• Research article review: Each review comes with a list of suggested edits.
• Poster board tutoring: We will teach you how to arrange a beautiful poster board.
• PowerPoint tutoring: We will teach you how to craft a compelling PowerPoint.
• Oral presentation tutoring: We will teach you how to become a great presenter!

Note: we cannot guarantee any particular outcome. Please refer to our Student Accomplishments to see past student results.

REGISTER:

To register, please reach out to hannah@thecuriocat.com. We do offer some discount based on your financial situation. Please inquire.

The post Advanced AI Research Program appeared first on The Curio Cat.

As a fun exercise, the CurioCat team competed in our first ever IEEE BigData Cup Challenge, entitled “Building Extraction”. The premise of this competition was to design a neural network to extract building footprints. Specifically, given a satellite image, the neural network must do three things: (a) detect individual buildings, (b) fit each building’s footprint to a polygon, and (c) list the vertices of every polygon. Below is a preview of a test image and our answer. Looks good so far!

What is Instance Segmentation?

In Computer Vision (CV), this task is known as instance segmentation. “Instance” indicates that each building is to be segmented individually. Semantic segmentation, on the other hand, treats objects of the same class as one entity, meaning that a cluster of buildings is segmented as a giant blob.

The best way to understand the task is to look at the training data. Our training data consisted of 9000 image pairs. The input is a satellite image and the output is the corresponding mask of building footprints. Below, we show an example of one image pair. On the right is satellite image of a Chicago neighborhood. On the left is the mask showing three classes: background (purple), building interior (green), and building outline (yellow). Eagle-eyed readers will note that the mask has some noticeable errors! This is because a human had to segment this image by hand!

Creating a Neural Network for Image Mapping

So, our task boils down to creating a neural network that can “map” the input image to the output image. This task is challenging because the test images consists of buildings in a variety of settings: (1) rural, (2) suburban, (3) urban, (4) commercial, (5) cityscape, (6) industrial, and (7) seaport. To be effective, a neural network must “generalize” well over all potential settings.

To do this, we created a neural network using a Resnet50 backbone. Resnet50 is a type of U-net architecture that is specifically designed for mapping image pairs. After our neural network predicts the output mask, we can easily detect and extract building footprints using the Python function cv2.findContours. Below are some more of our test results for different settings:

Why Did Our AI Hit a Ceiling??

So, how did our neural network perform overall? Our neural network was excellent in rural, suburban, and urban settings. However, our neural network was relatively weak in commercial, cityscape, industrial, and seaport areas. Why is this the case? Where did we go wrong?

Luckily, the answer is simple. Remember, a neural network can only go as far as its training data allows! So, if you want your neural network to learn seaports, you need to show it many seaports. Alas, due to time constraints, the training set we built was limited to certain types of areas. Specifically, we trained our neural network on the INRIA dataset for Tyrol Austria, Kitsap WA, and Chicago IL, which consists almost exclusively of rural, suburban, and urban areas. Therefore, it makes sense that our neural network would outperform in these areas.

Lessons Learned:

If we could do this BigData competition over, here is what we would change:

• We would have added more commercial, city, industrial, and seaport areas into our training set.
• We would have trained our neural network for more than 3 epochs. Research suggests 10-100 epochs.
• Our neural network was fooled by objects that look like buildings, including shipping containers, satellite dishes, grain silos, storage tanks, swimming pools, large air conditioning units, and even tennis courts. With this knowledge, we would have added images of these objects to inform our neural network of what not to segment.
• Our neural network was fooled by shadows. In images where the sun was low, peaked rooftops possessed different shades: the side facing toward the sun and the side facing away from the sun were lighter and darker, respectively. Instead of segmenting the building as one building, the neural network mistakenly “saw” two buildings. So, our training set should have included satellite images from different times of day.
• Our training dataset consisted of moderate resolution images. However, the testing set contained some grainy images. To strengthen our neural network’s ability to generalize, we could have injected Gaussian noise into our training set.
• Our training images were more zoomed-in than our testing images. So interpolating down our training images may have been helpful.
• For our multiclass classification, we gave our three classes, i.e., background, interior, and outline, weights of 0.25, 0.25, and 0.50, respectively. However, given that there are far fewer outline pixels, perhaps we should have given the outline an even greater weight.
• We would have experimented with different backbones. Instead of Resnet50, we could have tried DeepLab, Mask R-CNN, or YOLO.
• Finally, we could have tried a new technology like vision transformers.

The post IEEE BigData Cup Challenge – Lessons Learned! appeared first on The Curio Cat.

Our winter AI Research Program is now waitlisted – be sure to stay tuned for details about our upcoming program offerings and availabilities.

Our AI student research programs run on a seasonal quarterly system (like most schools) and the dates through next summer are as follows:

Fall 1-on-1 Program: Aug 8, 2024 – Nov 16, 2024 (0/8 spots open)

Winter 1-on-1 Program: Dec 1, 2024 – Feb 15, 2025 (0/8 spots open – waitlist available)

Spring 1-on-1 Program: Feb 23, 2025 – May 10, 2025 (4/8 spots open)

Summer Beginner AI Research program: June 15, 2025 – July 20, 2025

Summer Advanced AI Research program: June 14, 2025 – July 19, 2025

To sign up, or if you have any questions, please fill out our contact form here or email us at hannah@thecuriocat.com today!

Keep up with The Cat on Instagram or Facebook for updates and current events!

The post AI Research Program Winter 2024 – WAITLISTED! appeared first on The Curio Cat.

We joined over 5500 mathematicians and math enthusiasts in downtown San Francisco this year to attend the 2024 Joint Mathematics Meeting: the world’s largest math meeting! And for the second year in a row, our awesome students were accepted to present their research!

The conference was a whirlwind of dancing, food, Rubik’s Cubes, math art, presentations, and fascinating lectures – we even learned how chicken nuggets can teach us linear algebra LOL!

Check out our latest video to experience what it was like at JMM 2024 – the world’s LARGEST math meeting!

Or watch it on our YouTube channel here. To learn more about Hannah, our videographer, click here!

The post Joint Mathematics Meeting 2024 appeared first on The Curio Cat.

We were so excited to witness SpaceX’s latest rocket launch/flight over the Southern California desert last week! It just so happens that we’ve been building some AI rocket simulations for an upcoming student research project studying the effects of aerodynamics and various rocket designs.

Check out a little demonstration of some of our rocket A.I. sims as well as some awesome 4K footage of the SpaceX rocket flying over the desert!

Or watch it on our YouTube channel here. To learn more about our AI sim guru, Derek, click here!

The post AI Rocket Simulations appeared first on The Curio Cat.

Water is one of our most important resources – and also one of the most destructive forces of nature! In the video below, we give you a glimpse of an early water simulation being built by our students to predict and model potential damage from natural disasters.

Know any students interested in AI research? Click here to check out our AI Student Research Academy!

Or watch it on our YouTube channel here!

The post Water Simulation Sneak Peek!! appeared first on The Curio Cat.

In this Python tutorial, I will show you the second part of how to create a BlackJack command line game. This project will teach you coding strategy and help you become more comfortable with strings, arrays, and while loops. In a future video, we will repeat this BlackJack game using a Graphical User Interface (GUI) instead. So be sure to subscribe to stay tuned!

Or watch on our YouTube channel here. To learn more about the video creator, Derek, click here!

The post Python Tutorial: appeared first on The Curio Cat.

In this Python tutorial, I will show you how to create a BlackJack command line game. This project will teach you coding strategy and help you become more comfortable with strings, arrays, and while loops. In a future video, we will repeat this BlackJack game using a Graphical User Interface (GUI) instead. So stay tuned!

Or watch on our YouTube channel here. To learn more about Derek, the video creator, click here!

The post Python Tutorial: BlackJack Command Line Game appeared first on The Curio Cat.

Or watch it on our YouTube channel here. To learn more about Derek, the video creator, click here!

The post Python Lesson 12: Try Except appeared first on The Curio Cat.

Or watch on our YouTube channel here. To learn more about Derek, the video creator, click here!

The post Python Lesson 11: User Input appeared first on The Curio Cat.