Rekognition DetectModerationLabels 7.0 API Update Unreliable

0

Hi folks, Reaching out because I've noticed that the newest updates to DetectModerationLabels API has been very wrong for some of the content being reviewed. Some content is getting flagged with Explicit Sexual Activity at a confidence level over 95% that absolutely has none. Using github.com/aws/aws-sdk-go v1.51.0 Has anyone else seen this behavior?

Loren
asked a month ago135 views
1 Answer
0

Thank you for reaching out to the Rekognition team. We appreciate your feedback. For next steps, we recommend the following: (1) To improve the prediction accuracy immediately, you can try out our feature, Custom Moderation, which enables you to finetune our model with your own images for improved accuracy on your moderation use case. (2) We would appreciate if you can open a support case and provide a few images that are being flagged incorrectly. We are happy to answer any further questions you may have. Thank you for being an Amazon Rekognition customer.

AWS
answered 25 days ago
profile picture
EXPERT
reviewed 25 days ago

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions