FireLens

Igniting Awareness, Saving Lives

Comment

Inspiration

The project was inspired by the pressing need to raise awareness about preventable urban health risks, specifically house fires. We recognised that many people aren't fully informed about fire safety measures and the potential consequences of house fires. We aimed to create a powerful, engaging way to communicate this important issue, using visuals and narratives to make the information resonate with viewers.

What it does

Our project is an AI-generated video highlighting house fire dangers from multiple perspectives, including those of affected individuals, emergency responders, and healthcare professionals. The video is designed to educate and inform viewers on the risks associated with house fires, encouraging them to adopt preventive measures to protect themselves and their communities

How we built it

We built the project using a combination of Runway ML and CapCut tools. Jesse generated initial scenes using Runway ML, creating images of actors and converting these into videos that captured the emotional intensity of each scenario. Kofi handled prompt engineering and video editing, utilising CapCut’s AI-powered editing tools to refine the visuals and create voiceovers. ChatGPT assisted in generating a cohesive script, adding narrative structure and flow to the video.

Prompts runwayml

  1. Create an 18-second video of an old black organic farmer (using the provided image as the subject) looking directly at the camera with a mild yet slightly disappointed expression. Keep the camera focused on his face, capturing the subtle emotions in his eyes and expression. In the background, his farm environment gradually deteriorates over the course of the video. Lush crops and greenery behind him slowly wither, fade, and dry out, symbolizing environmental challenges or agricultural hardship. Use smooth, gradual transitions, ensuring the change in the background is noticeable but does not distract from the farmer’s solemn, steady gaze. The lighting remains soft and natural throughout, reinforcing a contemplative and emotional tone.

  2. Create a portrait of a white firefighter standing in front of a burning house. The camera should be tightly focused on his face, capturing his expression in vivid detail. He wears slightly crooked glasses, with ashes smeared across his tired, weathered face. His expression reflects a mixture of exhaustion, determination, and emotional weight from the situation. The glow of the fire casts warm highlights on his features, while faint smoke swirls in the background, creating a sense of urgency. Ensure the flames and house remain slightly blurred to keep the main focus on the firefighter’s intense gaze

  3. Create a 15-18 second video of a white firefighter standing still with a weary, determined expression. The camera stays tightly focused on his face, ensuring every detail—his slightly askew glasses and the ash smudges on his skin—remains visible. In the background, a house burns with visible flames and light smoke rising, but the fire stays subtly blurred, ensuring the focus remains on the firefighter's gaze. The flickering firelight softly illuminates his face, adding warmth and depth to the scene. The firefighter stays calm and resolute, looking directly into the camera throughout, reflecting quiet strength amidst the chaos

  4. Start with the firefighter standing in front of the burning house, in focus. The camera should rise slowly into an overhead, bird’s-eye view. Once it reaches the top, the camera should hover briefly over the house, allowing the flames, smoke, and roof to be clearly visible. After this pause, the camera should glide forward, soaring over the house as if passing by from above. Maintain a smooth, steady motion, simulating the natural flight of an eagle. As the camera moves past the house, let the fire and smoke remain visible, with embers flickering into the air. Ensure the lighting reflects the warm glow of the flames, and the motion stays cinematic and fluid, capturing the sense of flight while panning away from the house.

  5. Create an 18-second video of a female nurse standing outside a hospital, captured from the chest upwards. The camera stays tightly focused on her face, emphasizing her very sad and worried expression and the subtle emotions in her eyes. As the video progresses, a plain white and red ambulance pulls up behind her, clearly visible but not overpowering her presence. The background should show the hospital and the ambulance in focus, providing context to the scene without distracting from the nurse. The lighting is soft and natural, creating a poignant atmosphere that highlights her vulnerability amidst the urgency of the healthcare environment.

  6. write a script for morgan freeman to read regarding the issue of house fires and how they polute the environment. the first scene of the video contains the black farmer sitting in focus with a green environmental feeling background. the next scene is a sudden change to him in front of his farm with smoke coming from fires in the back and all his crops destroyed. the next scene is a fire fighter in front of a burning house. next scene is the same fire fighter in front of the house. next scene is a nurse in focus. the next scene has an ambulance pulling up behind the nurse. each scene is 10 seconds long

Challenges we ran into

One of the major challenges we faced was achieving accuracy in scene generation. Initial prompts often didn't produce visuals that matched our intended scenes, leading us to iteratively refine our prompts to better capture the emotions and scenarios we wanted to convey. This process taught us much about the importance of precise, prompt engineering in AI-generated media.

Accomplishments that we're proud of

We're proud of creating a high-quality, impactful video that communicates an essential urban health issue through AI-enhanced storytelling.

What we learned

We learned about prompt engineering throughout the project, especially how small prompt changes can dramatically alter AI output. We also became more proficient in working with AI-powered video editing and voiceover tools, expanding our multimedia storytelling skill set. The experience underscored the value of persistence and iteration in creative AI projects.

What's next for FireLens

We aim to integrate the video into educational campaigns, adapt it for social media, and expand this storytelling format to cover other urban health topics.

Built With

  • capcut
  • chatgpt
  • runwayml
Share this project:

Updates