The Amazon Rekognition service recently picked up three major updates, which make it easy to add deep learning computer vision analysis and capabilities to the cloud.
First of the three features is recognition of text like street names in images, the second being real time face detection, and finally, we have improved detection of faces in a crowd.
All three have been designed to make to machine learning more accessible for AWS customers, and offering them tools to implement in their products. Ultimately, making the cloud platform a more appealing option than its rivals.
All four major cloud giants — Amazon, Microsoft, Google, and IBM — are infusing their products and platforms with artificial intelligence and machine learning capabilities.
Getting back to these latest updates, the first feature that AWS has introduced is Rekognition Text In Image, which as the name suggests enables cloud customers the ability to extract textual content from real world images like street signs, license plates, and even the news.
Next up we have real time face identification across tens of millions of faces, with the improved face search offering a reduction in search latency to the tune of 5 to 10 times. It is also simultaneously able to store 10 to 20 times more faces.
These capabilities could come in really handy for law enforcements agencies.
And finally, the service can recognize up to 100 faces in crowded photos, and this advancement in detection will enable companies and businesses to accurately receive information about the demographics, sentiments, and emotions of all the faces detected in a single photo.
This is a significant upgrade from the previous limit of 15 faces.
All these new additions show the investments Amazon is making in AI and ML, and lay the groundwork for more such impressive enhancements for the already impressive Rekognition service.