ML Newsnew | past | best | rss | submit about
Acoustic pornography recognition using CNN and bag of refinements (arxiv.org)
6 points by PaulHoule on Nov 14, 2022 | hide | past | pdf | 8 comments on HN

In plain words: A neural network listens to audio and decides whether a recording is pornographic, so it can be filtered to protect children. Tested on a large audio collection, the base model reached 92.46% accuracy, and stacking all its tweaks raised that to 97.19%.

Abstract · Acoustic Pornography Recognition Using Convolutional Neural Networks and Bag of Refinements

A large number of pornographic audios publicly available on the Internet seriously threaten the mental and physical health of children, but these audios are rarely detected and filtered. In this paper, we firstly propose a convolutional neural networks (CNN) based model for acoustic pornography recognition. Then, we research a collection of refinements and verify their effectiveness through ablation studies. Finally, we stack all refinements together to verify whether they can further improve the accuracy of the model. Experimental results on our newly-collected large dataset consisting of 224127 pornographic audios and 274206 normal samples demonstrate the effectiveness of our proposed model and these refinements. Specifically, the proposed model achieves an accuracy of 92.46% and the accuracy is further improved to 97.19% when all refinements are combined.

Lifeng Zhou, Kaifeng Wei, Yuke Li, Yiya Hao, Weiqiang Yang, Haoqi Zhu
arXiv:2211.05983 · cs.SD, eess.AS · submitted Nov 11, 2022
abstract · pdf · html

add comment on HN

sqeaky on Nov 14, 2022 [–]
> A large number of pornographic audios publicly available on the Internet seriously threaten the mental and physical health of children

First sentence. This is either a great humanitarian scouring for the victims of child porn. Or it is a useless puritanical crusade against porn.

I started reading the paper and this is just about porn filtering, pretty much useless perhaps even harmful.

I am sure using CNNs to do this is innovative, but if you are going to start from a lie why should I trust your research?

reply
ljclifford on Nov 14, 2022 | parent [–]
I'm unfamiliar with 'pornography recognition' as an established task in ML research (lol), but for what it's worth, it's not an innovate use of CNNs for audio classification. You can essentially turn any audio classification task into an image problem (raw audio into features like spectrograms/MFCCs). Which people have been doing since forever (by which I mean a number of years now).
reply
PaulHoule on Nov 14, 2022 | parent [–]
The question of it in my mind is if the available features are structurally adequate.

Porn videos frequently have a soundtrack with distinctive moaning and that's one thing. If the problem was distinguishing audio recordings of somebody talking about erotic or not-erotic subjects, that's something entirely different.

reply
sqeaky on Nov 14, 2022 | parent [–]
> 'pornography recognition' as an established task in ML research (lol),

Exactly my thought process. Maybe I should learn modern ML and apply my, ahem expertise, and apply for some grants.

reply
mosseater on Nov 14, 2022 | parent [–]
Porn filtering is useless or harmful?

It's like telling me you don't have kids without telling me you don't have kids xD.

I have my 8 year old sons internet pretty much locked down at this point, but people still end up skirting around the rules of sites like youtube when publishing content towards kids. Automatic filtering is one of the only ways to prevent newly published content geared towards children from being malicious. The alternative, human curation, even if done by method of reporting a video, is time consuming and expensive.

reply
PaulHoule on Nov 14, 2022 | parent [–]
Whether you like porn or hate porn, classifying porn is useful. There is a lot of it around and it causes a lot of practical problems if you are hosting public content if you're not in control of it.
reply
hkxer on Nov 14, 2022 | parent [–]
Porn has been proven to have severe detrimental physical and mental effects, including: erectile dysfunction, depression, changing perceptions of sex, etc.

Feel free to google them (going on google scholar and typing in “pornography negative impacts” is a good start), or I can cite a laundry list of studies saying the same thing.

reply
withinboredom on Nov 14, 2022 | parent [–]
I mean, if you’re going to bias the search results, you might as well also search for the opposite before posting a comment…
reply
About | RSS | RSS (all) | HN arXiv