Yahoo has designed an NSFW detection bot that uses deep learning to assess images and give them scores based on how appropriate they are.
NSFW, an acronym for "not safe for work," is a common label used on the internet to demarcate inappropriate content. Typically, this means that the content is either excessively gory, contains swearing, or is pornographic. Yahoo's current deep learning algorithm only offers NSFW detection for pornographic images, and, further, it can only detect photographs (as opposed to sketches or cartoons, for instance.) Nonetheless, it is still a far more efficient detection system than manually checking every image in a given space.
The NSFW scores in the Yahoo system are graded from zero to one, with zero being completely safe and one being completely inappropriate.
NSFW Detection Algorithms
Yahoo Uses Deep Learning to Assign NSFW Scores to Images
Trend Themes
1. Deep Learning NSFW Detection - Opportunity for companies to develop deep learning algorithms for NSFW detection, which can improve efficiency and accuracy of content moderation.
2. Automated Content Moderation - Companies can invest in automated content moderation systems using NSFW detection algorithms, to reduce manual efforts and improve user experience.
3. NSFW Detection for Various Content Forms - Companies can explore the opportunity to develop NSFW detection algorithms for videos, cartoons, and other forms of media, to broaden the type of content moderated and improve online safety.
Industry Implications
1. Social Media - NSFW detection algorithms can be integrated into social media platforms for effective content moderation, to improve user experience and reduce the spread of inappropriate content.
2. Online Marketplaces - E-commerce sites and marketplaces can integrate NSFW detection systems to ensure the safety of product listings and prevent the sale of inappropriate items, improving user trust and satisfaction.
3. Ad Tech - Ad tech companies can leverage NSFW detection algorithms to ensure brand safety and prevent the display of inappropriate ads, reducing ad fraud and improving user experience.