Inspiration
Use and learning of AI techniques that work with images
What it does
It finds if the images are similar or not. Similarity is defined by comparing the clothing of the images by the appearance such the color.
How we built it
We used existing image segmentation AI models to clean the image of noise that we thought it wouldn't help at deciding if the images were similar o not (ex: background, model, etc). We also used image feature extraction and models that check similarity of images once they've been cleared of noise.
Challenges we ran into
- Defining what similarity is.
- Defining noise in the images and removing it as intended.
- Looking for ways to lower the execution cost.
Accomplishments that we're proud of
- We found ways to removing the noise of the image with a good quality, even though it came with a high cost.
What we learned
How to use AI image models.
What's next for Inditex Image Comparison
- Look for applications in the market as helping find a client new clothes to buy.
- Upgrade the efficiency of the algorithms.
Built With
- gradio
- python
Log in or sign up for Devpost to join the conversation.