sd-webui-nsfw-filter algorithms of the filter might be too strict

i had tried with a few checkpoints, as shown above. Not sure if it's possible to have a weight option in order to adjust filtering rules?

Oct 22 '23 14:10 sdwcn

Sure, I'll add some thresholds and options for filtering (maybe, within 9 hours)

Oct 22 '23 18:10 aria1th

Sure, I'll add some thresholds and options for filtering (maybe, within 9 hours)

how's it going?

Oct 25 '23 11:10 sdwcn

@sdwcn I tried some more setups, but seems like it has weird bias on white dress... The first plan was implementing labels to blur, but maybe it should have some very custom setups (such as thresholding by probability) Today, I'll release with some basic and custom setups, but I think model was roughly underfitted for border-images, resulting 'some' exposure to be censored

Oct 26 '23 00:10 aria1th

noted, I'd like give a try,

Oct 27 '23 06:10 sdwcn

how shall I understand the diff between safe questionable & explicit?

levels_to_idx = {
        'safe': [1,3,4],
        'questionable': [1,3],
        'explicit': []
    }

according to tags in comments, ['drawings', 'hentai', 'neutral', 'porn', 'sexy'] does qustionable equals [hentai + porn]?

Oct 27 '23 06:10 sdwcn

Yes, safe = all (hentai porn sexy) questionable = hentai+porn explicit = none

Oct 27 '23 09:10 aria1th

switched to "questionable", it still bias on white dresses, it sounds good if a probablity setting is enabled. cheers

Oct 30 '23 02:10 sdwcn

@sdwcn I can offer the probability option but its hyperparameter may require tremendous amout of test 😭

Oct 30 '23 03:10 aria1th