Vision COP: Image Search Engine (Image.search)

• 

  the first picture is input image, and the rest is the results.

Inspiration

Advances in AI especially computer vision make a lot of benefits. one of them is the image search engine, this makes the search not only text-based. In addition, image search engines can be used for product based recommendations for example in e-commerce.

What it does

It can find similar image in database from user input image.

How we built it

Using ML to extract features from the input image and measures the distance of the features with the features of the image database. So that it can find similar image in database from user input image.

Challenges we ran into

-Got many errors during model deployment in the cloud, so it's still not deployed in the cloud.

Accomplishments that we're proud of

-Using AI to do image search

What we learned

Tensorflow, Azure Machine Learning

What's next for Vision COP: Image Search Engine (Image.search)

Since the dataset used only CBIR dataset, which contains animal images, the model can't find similar images if it used for searching other objects. the more database images will be beneficial so it can used widely.

Built With

• azure
• matplotlib
• numpy
• python
• tensorflow