Amazon cloud business is launching artificial intelligence tools, and the Rekognition Video tool is one of them. Rekognition Video said to have the ability to track people in video feeds and has sure advantages over other video recognition tools that were launched by Microsoft and Google.
Amazon Rekognition service launched in 2016. It is a cloud service and provides image recognition and analysis by using deep learning. Amazon Rekognition Image enables building and integrating object, facial recognition, celebrity recognition, text recognition and image moderation in systems and applications.
The service was released initially focusing on providing scalable and automated analysis and classification of images. Some additional features now offered in the service.
Table of Contents
What is Amazon Rekognition and what is Deep Learning?
Amazon Rekognition service makes adding visual analysis to the applications. Powerful applications can build by using this service to search, organize and verify images in large numbers. Rekognition image tools detect scenes, faces and also extract text.
It identifies celebrities and also recognizes inappropriate content in pictures. Face comparison and searching are also possible. Rekognition video was developed based on Deep Learning technology.
Rekognition Video is another tool that tracks people, activities, objects, celebrities and inappropriate content in videos in live video streams also. The tool can monitor the persons even when the faces are not visible.
Deep Learning is a field of Machine Learning in Artificial Intelligence and based on information processing and communication in the brain. Amazon Rekognition is part of Amazon Artificial Intelligence service, and this service uses deep learning to search and understand images, etc.
Real-Time Video analysis with Rekognition Video
Amazon Rekognition Video is one of the pre-trained machine learning tools of Amazon’s AWS division. Batch uploads and real-time video streams can make use of this new service. Getting the information of the things in the video is possible. This service also supports person detection and recognition. It can filter inappropriate content and can track people in the video.
Recognizing objects and people in images is more comfortable than in the videos. The focus of several in the recent years was on video recognition and the services like Rekognition developed. It can generate metadata from video files.
Key Features of Rekognition Video
Amazon Recognition Video can analyze and detect faces in live video streams provided from Amazon Kinesis Video feature can be used for public safety applications and can help in crime prevention. Rekognition Video can also analyze video data stored on Amazon S3.
- Person identification and tracking are possible in videos in real time with Rekognition Video. Track person API detects the persons even when the camera is in motion or when the face of the person blocked. It helps in retail applications to generate customer insights and in public safety applications, this helps in monitoring and investigating specific persons.
- Face Recognition is another feature of Rekognition Video. The user can create face collection by using CreateCollection API. Rekognition searches this face collection for similar faces in the video and returns the results. Face Recognition helps in crime prevention as it becomes possible to identify persons of interest among a group of millions of faces.
- The Facial analysis is another feature of Rekognition Video. This locates faces that appear in the video and reports its emotions. Landmark points such as eyes, nose, and two corners of the mouth returned.
- Rekognition Video also detects suggestive and inappropriate content in the videos. Based on application requirements, the videos can filtered, and right moderation level can apply to the audience.
- Another exciting feature of Rekognition Video is the celebrity recognition. This tool can detect when a celebrity appears in the video.
Google Cloud Vision Vs. Amazon Rekognition
These two are the most useful services for video recognition and some technical aspects can differentiate the two. Some high-level features and the support of the platforms compared below.
- Object detection supported by Google cloud vision and Amazon Rekognition.
- Face detection feature is available in both the services.
- Face comparison and Face search features are available in Amazon Rekognition.
- Amazon Rekognition supports JPG and PNG formats and Google Cloud vision supports most other image formats.
- Object detection functionality is similar to both the services.
- Amazon Rekognition is better at detecting individual objects such as humans, glasses, etc. compared to Google Cloud Vision.
Amazon Rekognition Video can recognize faces, activities, objects, celebrities and inappropriate content in the videos stored in Amazon S3 and across live stream videos. The API powered by computer vision models that are pre-trained to detects objects, activities in thousands of numbers. The tools also can tag specific sections of the video with labels. It can identify activities such as running, walking, swimming, etc.
Other tools launched by Amazon are the Amazon Transcribe that can transcribe the audio recording of people into text files. Amazon Comprehend is another service that picks up positive/negative sentiments, people, places and phrases in the text. Amazon Translate is another useful service that and translate text from one language to another.
This service already offered by Google for years. Amazon says they aim to solve the problem of Artificial intelligence accessibility for developers and scientists. Andy Jassy, CEO of AWS was speaking to a gathering of over forty thousand people and announced the new AI services.
Amazon Rekognition service makes adding visual analysis to the applications. Rekognition Video help people to identify the persons to prevent the crimes and increse the public safety. Know more about Rekognition Video make use of the above guide.