Some very cool news to report from Clarifai HQ today!
We have two brand-spanking new models ready for testing and we'd love to hear your feedback on them:
The 'General Embedding' model analyzes images and returns numerical vectors that represent the input images in a 1024-dimensional space. The vector representation is computed by using our 'General' model, and is the driving force behind our Visual Search platform.
Embeddings can be used for filtering, indexing, ranking, and organizing images according to visual similarity.
Most common use cases would be:
- Visual search (if you want to build your own, and don’t want us to index your images)
- Research experiments using the embeddings
- Perceptual hashing
- Multi-modal classification system
Check it out here!
Not to be outdone, in corner #2......
The 'Focus' model analyzes an image and returns:
- The overall focus value (0-1) that represents the probability that there is an in-focus region within the image
- A bounding box and focus density for every in-focus region within the image.
This model is great for anyone building an app that relies on filtering out blurry images, and can include any photographs that are taken in a non-still environment -- e.g. sports, aerial photographs (drones), event photography, and more.
Try it out here!
Very curious to hear everyone's feedback on them!