Questions tagged with Amazon Rekognition

I've been on this for about 2 days now and it's extremely frustrating. What I need is to use detectFaces / compareFaces and I want to supply the image files by specifying "Bytes". But after trying all kinds of different encodings and formats, I always end up with an InvalidImageFormatException. In order to make sure that was actually the problem, I decided to make things simpler and use images from an s3 bucket. Now, when I detect-faces from the command line, it works! But when I run the same exact thing with my node app, by specifying the bucket file: ``` { Image: { S3Object: { Bucket: "bucket.test", Name: "test.jpg" }, }, Attributes: ['ALL'] } ``` I get an InvalidS3ObjectException .... And I am quite stumped. Why is it so difficult to provide inputs to Rekognition? Azure just lets you specify image urls and it works no problem. AWS gives vague error responses and it's unclear what the real problem is.lg...

I want to use rekognition to create an identity verification workflow such as https://docs.aws.amazon.com/rekognition/latest/dg/identity-verification-tutorial.html. I need to validate a selfie against an ID card. My problem is how I can recognize that the selfie photo is not a "fake/spoof" image, ex: I use an already taken image as a selfie or use the ID card in both images. Specifically, Is there a way to recognize Presentation Attack Detection (PAD)?lg...

I want to use python to connect with the AWS Rekognition to get the detect label results, but I have been trying this for a long time, and whenever I want to use python to connect to my project which I chose the "bounding box" one in Rekognition, and it shows "botocore.errorfactory.InvalidImageFormatException: An error occurred (InvalidImageFormatException) when calling the DetectCustomLabels operation: Request has invalid image format". I have read the developer guide and I know I use the right format which is "PNG". I would wonder if there are something that I miss? Thank you very much.lg...

We are trying to get customer's face image details using AWS rekognition's `detectface `API. We are passing the document in bytes as input, but this is highly restricted data as it contains Unique Customer Information. `List<FaceDetail> detectFaces(@NonNull final byte[] bytes) ` We know service supports over HTTPS and encrypted with TLS. Is this client to server channel safe from any external intrusion?? OR Does this AWS rekognition provide any encrypted input data transmission support for e.g. client side encryption at rest ?lg...

I trained a model in rekognition and run in aws cli. But ı call the method return empty list like this: "CustomLabels": [] How can I solve ıt. Thank you.lg...

Hi everyone, Looking at the Rekognition [Guidelines and quotas](https://docs.aws.amazon.com/rekognition/latest/dg/limits.html) I don't see any mention of how many Collections you can have in an account. Would there be any problem in having millions of collections? Thanks for your help.lg...

Has anyone had the need to custom build and containerize of our cognitive services so it can run on premise? ( Speech, vision, image etc)lg...

When using Rekognition, if you previously saved 1 million images in a Collection and do a SearchFaceByImage of 1 new image against those 1 million images, do you need to pay for 1 million API calls?lg...

This image is being identified as explicit nudity via the Rekognition detectModerationLabels API Is there a better place to report issues like this?  https://ic-cdn.flipboard.com/meredithcorp.io/ea9f9f3bcd15ae62bfcbdfe0a12879b8b8ee3287/_large.jpeg ``` "moderationLabels": [ { "name": "Explicit Nudity", "parentName": "", "confidence": 92.17741 }, { "name": "Illustrated Explicit Nudity", "parentName": "Explicit Nudity", "confidence": 92.17741 } ] ```lg...

I am trying to use the AWS Rekognition Service ( Java SDK ) & adding faces to an existing collection. But I am getting following issue Error :"Request has invalid image format (Service: Rekognition, Status Code: 400, Request ID: 4adc5ed9-3484-420a-bb78-34aebe6d8bbb)" As per the documentation from the below URL for Java, no need to do base64 for images. https://docs.aws.amazon.com/rekognition/latest/dg/images-information.html I am not sure what I am missing here. Can anyone assist here? ************** public class AddFacesToCollection { public static final String collectionId = "vowCollection"; public static final String bucket = "vowvideo"; public static final String photo = "sachin.jpeg"; static LoggerContext loggerContext = (LoggerContext) LoggerFactory.getILoggerFactory(); static { loggerContext.getLogger("org.apache.http").setLevel(Level.INFO); loggerContext.getLogger("com.amazonaws").setLevel(Level.INFO); } public static void main(String[] args) throws Exception { AmazonRekognition rekognitionClient = AmazonRekognitionClientBuilder.defaultClient(); Image image = new Image() .withS3Object(new S3Object() .withBucket(bucket) .withName(photo)); IndexFacesRequest indexFacesRequest = new IndexFacesRequest() .withImage(image) .withQualityFilter(QualityFilter.AUTO) .withMaxFaces(1) .withCollectionId(collectionId) .withExternalImageId(photo) .withDetectionAttributes("DEFAULT"); IndexFacesResult indexFacesResult = rekognitionClient.indexFaces(indexFacesRequest); System.out.println("Results for " + photo); System.out.println("Faces indexed:"); List<FaceRecord> faceRecords = indexFacesResult.getFaceRecords(); for (FaceRecord faceRecord : faceRecords) { System.out.println(" Face ID: " + faceRecord.getFace().getFaceId()); System.out.println(" Location:" + faceRecord.getFaceDetail().getBoundingBox().toString()); } List<UnindexedFace> unindexedFaces = indexFacesResult.getUnindexedFaces(); System.out.println("Faces not indexed:"); for (UnindexedFace unindexedFace : unindexedFaces) { System.out.println(" Location:" + unindexedFace.getFaceDetail().getBoundingBox().toString()); System.out.println(" Reasons:"); for (String reason : unindexedFace.getReasons()) { System.out.println(" " + reason); } } } -Snehadeep.lg...

What’s your experience with content writing and Creative writilg...

I take individual images and then composite them into essential collages. I would like to feed Rekognition the raw unedited images, and then have Rekognition find and label them in the collage image. Is this possible?lg...