General

 

What is Kairos?

Kairos is an artificial intelligence company that specializes in face recognition, making it easier than ever for businesses to transform how they interact with people. Through computer vision and machine learning Kairos can recognize faces in videos, photos, and the real-world - Learn more about Kairos.

 

What features does Kairos offer?

Face recognition, emotion, and demographic analysis - See all features.

 

Can you provide case studies?

All kinds of businesses use Kairos to add face analysis in their products. We’re trusted by global media agencies, innovative start-ups, and some of the world’s largest organizations.

We have over 10,000 developers on our platform processing millions of faces each month - See how customers use Kairos.

 

How accurate is Kairos face recognition?

We strive for 100% accurate results. In reality face recognition accuracy largely depends on the quality of submitted images/video and the settings you determine through our API. Our algorithms can compensate for non-ideal conditions such as uneven lighting or side poses.

Like any technology, the more controlled the environment, the better results you get. If you are not getting the outcomes you expect, we can work with you to improve your results - Learn more about Best Practices for Face Recognition.

 

Do you have a Demo?

Yes - Try our face analysis demo.

If you’d like to showcase Kairos at your company or event - This email address is being protected from spambots. You need JavaScript enabled to view it..

 

How do I use Kairos?

Integrate with our easy to code API - Choose a plan to get started.

 

Can I use Kairos for people counting?

Face recognition isn’t really designed for people counting, although you can still use it to get a sense of how many people come in and out of frame. You would have to consider your camera positioning to make sure you get as many people’s faces as possible - Learn more about Best Practices for Face Recognition.

 

Is there a limit to how many faces can be stored in a gallery?

No. You can store an unlimited number of faces in a gallery.

 

What does the ‘Human Analytics’ thing mean?

We call our features ‘Human Analytics’. You can use them to gather unique, real-time insights about people as they interact with your company, brand or product. Even though we’re focused on face analysis right now, in the future we will be expanding our products with other ways of measuring people - Stay tuned!

 
 

Security & Privacy

 

Is the Kairos Human Analytics API secure?

Yes. We are hosted at Amazon Web Services (AWS) and are a Standard Technology Partner of AWS. We offer 2048 bit SSL encryption for all data in transit and our backups - For more information on AWS, visit their compliance page.

All of the transactions you submit is through a user-generated API key. The API key you generate authorizes the engine to perform tasks on your behalf.

Payments are processed through our partner Braintree. We never see your credit card details.

 

Is my information private?

Yes. We do not use any personally identifiable information and any face template data we create cannot be reverse engineered. Your image classification in the API only needs to make sense to you. That means you can use anonymous or tokenized identifiers instead of personal data such as employee numbers.

 

What encryption do you use?

If you use our HTTPS API your data is encrypted in transit with 2048 bit SSL. We use encrypted storage volumes and our backups are encrypted as well.

 

Where is my data stored?

At Amazon Web Services’ secure cloud at multiple locations in the United States. If your company is located outside the US and you wish to host your data at your native country, we can host our infrastructure at other Amazon locations throughout the world such as Brazil, Germany, Ireland, Singapore, etc. - Contact us to learn more.

 
 

Human Analytics API

 

How does the Kairos Human Analytics API work?

You submit images or video into our API. We find all the faces and return detailed facial information back in the form of code objects (JSON). These results include feature points, emotion, age, gender, ethnicity, and a ton of other useful data about the human face - See all features.

You can also create galleries of faces from your images and videos to be used later for face verification or identification (this is what a lot of people call ‘face recognition’).

All of the transactions you submit is through a user-generated API key. The API key you generate authorizes the engine to perform tasks on your behalf.

 

What counts as an API transaction/call?

An API 'call’ or 'transaction' is an HTTP request sent across the Internet to our API.

E.g., A ‘call’ would be enrolling a face into a gallery to later be recognized, or a request sending us a photo with faces in to be detected. Requests to view galleries, remove faces, list galleries, are also a 'call'.

 

How fast is the Kairos Human Analytics API?

Fast. We typically match faces or enroll images in about a second. This speed can be affected by the size of the image, the number of faces in the image and your use case.

We’ve found that low resolution images work just as well as high resolution images as long as there are at least 75 pixels between the person’s eyes. The smaller the image the faster it is able to be processed. We can process images as small as 12 KB with great success.

For example, when you submit images we automatically compress them, and convert them to grayscale in order to speed up the process and save you bandwidth.

We do offer an offline version, which is faster because it doesn't need to send data across the Internet. However, we do recommend customers use the API to create their proof-of-concepts - Test for free before you scale.

 

What files does the Kairos Human Analytics API support?

Image files can be BMP, JPG, or PNG. We currently do not support GIF.

Video files can be AVI, FLV, MJPEG, MOV, MP4, MPEG, WEBM, or WMV.

PRO TIP
We recommend encoding your video with metadata that contains the video length, and using a frame rate between 20-25 fps.

 

What are ‘thresholds’, and how do they affect face recognition accuracy?

When you match faces against all the enrolled faces in your gallery, Kairos returns a confidence score between 0 and 1. The higher the number, the more confident the system is about the match.

Depending on your needs, you can set a threshold number between 0 and 1 where any value above that threshold would be considered a match and any value below that threshold would be considered a no match. For example, if you set your threshold at .82, any match with a score that is equal to or greater than .82 would be considered a match.

 

What types of cameras work with the Kairos Human Analytics API?

We’ve tested our API on a multitude of devices, from embedded smartphone cameras to webcams to DSLRs, and get equally good results from both ends of the spectrum.

The accuracy of your results mostly depends on the quality of your enrollment and recognition photos. It works best when there are at least 75 pixels between the person’s eyes - Learn more about Best Practices for Face Recognition.