AI ScreamScape

Inspiration

We both love haunted houses, but don't have consistent experiences with them since sometimes they have characteristics that we are scared of and at other times don't. Thus, we created a general problem statement and used Murphy's Law to inspire our project (as seen below). "What you dread the most tends to find its way, amplifying the power of your anxieties and validating your darkest apprehensions." - Murphy's Law Problem Statement: Many people enjoy the thrill of experiencing a haunted house, but finding one that caters specifically to an individual's fears can be challenging. Additionally, the level of fear experienced can vary greatly from person to person. There is a need for a personalized and immersive haunted house experience that can generate virtual reality (VR) or video content based on the user's specific fears and continuously adapt to their fear levels in real time.

What it does

Solution: To address this problem, we propose the development of a Haunted House AI system that generates a VR/video of a haunted house tailored to the user's fears and measures their heartbeat and breathing rate to reach the optimal point of fear. This system will provide a highly customized and interactive experience, maximizing the thrill and fear factor for each individual. Role of AI: The AI component of the Haunted House system plays a crucial role in creating the personalized experience. It will utilize machine learning algorithms to analyze user preferences, fears, and physiological responses to generate an immersive VR/video haunted house experience. The AI will continuously monitor the user's heartbeat and breathing rate through wearable sensors and use the data to dynamically adjust the level of fear-inducing elements within the experience. By leveraging AI, the system can adapt and optimize the fear factor in real-time, ensuring an engaging and personalized haunted house experience. Prototype Description: User Interface: The prototype includes a user interface where users can input their fears and preferences, providing information about the elements they find particularly scary (e.g. spiders, ghosts, dark environments). Fear Analysis: The AI component of the system analyzes the user's input and generates a fear profile. This profile categorizes the user's fears and assigns appropriate weights to each fear based on their intensity. Content Generation: Using the fear profile, the AI generates VR/video content of a haunted house. It combines pre-existing assets (e.g. haunted house environments, scary creatures) with user-specific elements that match the user's fears. The generated content is designed to elicit a range of fear responses. Inputs Monitoring: The prototype integrates wearable sensors that monitor the user's heartbeat, breathing rate, and other sensory inputs in real-time. The sensor data is fed into the AI system for analysis. Fear Optimization: The AI analyzes the user's data and adjusts the intensity of fear-inducing elements within the VR/video experience accordingly. It aims to reach an optimal point of fear for the user, maximizing their thrill while ensuring it remains within manageable levels. Real-time Feedback: The prototype provides real-time feedback to the user, indicating the system's analysis of their fear level and how it adjusts the experience based on their inputs. This feedback helps create a sense of engagement and immersion.

How we built it

We used inputs/outputs to create our idea User Preferences: The user can pick what they are afraid of (if they already know it) and the AI will start off and focus on those visuals. Otherwise, AI will use sensory detection methods to figure out the user’s fear/s.  Breathing Rate Normal breathing rate ranges from 12-16 bpm. When this is exceeded, it is a sign of fear. The AI registers this for users.  Heart rate Normal heart rate ranges from 60-100 bpm. When this is exceeded, it is a sign of fear. The AI registers this for users. 

Output: VR Visual

For our prototype, we created a Figma App design that allows users to choose what they are scared of. We are also using an Ai that generates film for our prototype, and we have physical models that are meant to model (but not function as) of the heart rate and breathing sensors and also of the physical VR.

Challenges we ran into

We wanted to find a project idea that didn't exist, and our group struggled to find that - but we both managed to create this ideathon as it is a product that doesn't exist.

Accomplishments that we're proud of

All the growth we have showcased through the conference (we were both here all three days), our ability to learn from the workshop and put it into our model (detailed below), and figuring out ways to create prototypes creatively to simulate what our product intends to do.

What we learned

Machine learning + INPUT/OUTPUT (PYTHON) How an AI changes when data/feedback is inputted In our product, we use inputs (heartbeat + breathing rate) from the user as the VR is working to make the machine better. We learnt this idea from the python workshops, as well as overall seminars.

Figma App design software skill Our app product uses Figma in our prototype. Collaboration/Idea Generation

Collaboration/Idea Generation We learned how to work with different people throughout this conference to improve our ideas. We also learned how to take the initiative and think out of the box, skills we used in our product as Haunted House AI doesn’t currently exist.

OPEN AI How to effectively use AI tools like ChatGPT We used this to get feedback on our language and verify that this product is unique.

What's next for AI ScreamScape

As we both develop our coding skills, we want to be able to code visuals (which are different from the coding languages we know and our skillset) to better manifest our product.

Built With

• aigeneratedvideos
• figma
• slides