This test is about tagging travel images. Hotels, pools, restaurants, out door sports, riverside walks, etc, etc... Everything that has to do with traveling.
First of all thanks to Goberoi Blog for using his python code to do the comparison of the Microsoft, Google, IBM, Clarifai and Cliqorange image recognition API.
I don’t want to hold you up if you want to go straight to the results. To get insight in this comparison of Image Recognition API’s you need to know the outline.
- This test is about travel images in the widest sense.
- We could not compare the “aesthetic score” as we are the only one offering this.
- The dataset used contains hotel related images and travel activity images. This covers all images that you will find on sites like Booking.com, Kayak, Airbnb and Expedia.
The general tags are all well trained by Google, Microsoft and all the other big companies. But specific travel “experiences” tagging is another cup of thea.
Together with the University of Amsterdam we have created an API to find specific travel tags so you can create finger licking landing pages!
We have a “general model” that finds 2000 tags and we have an “activities and experiences” model that can find around 120 tags. Check out our demo site Yellowroom.ai to see how to create the landing pages.
Room with city view
Room with sea view
Dinner on the beach
Room with jungle view
Luxury brand shopping
Fine art museums
Beaches for kids
Walking with pets
family with children
getting a face mask
Kiteboarding >change name in kitesurfing
rose leaves on the bed
Spa and relaxation
So In theory you can create landing pages of every tag posted above here. Check out our demo site Yellowroom to see an example.