Testing Sora and Make Sense AI for Mask Segmentation on Low-Quality Building Images: Amazing Results

The Beginning: A Low-Quality Building Image

I found an interesting image on the Aurecon Group website. The image showed tall buildings, but the quality was poor. The pixels were visible, details were blurry, and colors looked faded. Usually, I would think such images are too difficult to work with.

The Challenge: Mask Segmentation with Sora

I decided to try something new. I used the Sora system with Make Sense Alpha tool to do mask segmentation on the image. Mask segmentation is a technique that separates different parts of an image with high accuracy. For example, it can separate the sky, buildings, and towers individually.

The Result: An Amazing Surprise

What happened was truly incredible!

Sora did not just identify the buildings. It also:

• Distinguished each tower separately, even when they overlapped

• Separated the sky from buildings with amazing accuracy

• Recognized small details that I could not see clearly

• Created clean and sharp masks despite the blurry original image

What I Learned

This experience taught me something important: poor image quality does not always limit AI power.

Sora could "see" what we cannot see with our eyes. It could extract precise information from data that seemed limited. This is a perfect example of how technology can go beyond what we think is possible.

Conclusion

Working with Sora and Make Sense Alpha showed me the real power of modern AI tools. Even when starting with a low-quality image, the results can be surprisingly good. This opens new possibilities for image processing and analysis.