Artificial Intelligence Governance Professional AIGP · Free Practice Question Medium
Question 5
A social media platform adopts a deep learning approach to detect abusive content in multiple languages and formats. Users report that videos with audio are flagged less often than text or images. Which modification would MOST likely improve detection of abusive content in videos?
- A Apply the same text-based model to all formats
- B Switch from deep learning to rule-based keyword filters
- C Integrate multimodal neural networks capable of processing both audio and visual inputs
- D Reduce the size of the training dataset to include only videos
Reveal correct answer
Correct answer: C
Explanation
Multimodal deep learning models are best suited to jointly process and analyze varied data sources, such as audio-visual content.A. Models specialized for one modality may miss important signals in others.
B. Keyword filters lack the flexibility and sophistication for nuanced, multi-format content.
C. Multimodal networks handle multiple data types simultaneously, improving detection accuracy in complex content like videos.
D. Less data typically reduces performance and generalizability.
Discussion
Think the marked answer is wrong, or have a better explanation? Share it below — comments appear after review.
You must be logged in to post a comment.
