Which service can detect objects, scenes, faces, and inappropriate content in videos?

Amazon Rekognition Video is the service specifically designed for analyzing video content to detect objects, scenes, faces, and various types of inappropriate content. It leverages deep learning technology to provide precise and nuanced insight into the visual elements within video frames. This encompasses recognizing individuals, identifying emotions, detecting specific actions, and categorizing scenes, all of which are essential for a variety of applications such as content moderation, surveillance, and media analytics.

In contrast, Amazon Polly is a text-to-speech service that converts written text into natural-sounding speech, making it unrelated to video analysis. Amazon Comprehend is focused on natural language processing and can extract insights from text, such as sentiment analysis and entity recognition, but does not deal with visual content. Amazon Transcribe is designed for automatic speech recognition, converting spoken language into written text, which also does not include the capability to analyze the visual aspects of videos.

Thus, Amazon Rekognition Video stands out as the appropriate choice for detecting and analyzing visual content in videos.