One image of her holding an ice cream was turned into a sexualised video.
In another case, she was pictured with a cartoon body in a bikini, alongside posts mocking her.
“What were they hoping to achieve other than humiliate me?” she added.
Salt said 18 months ago she started to receive nasty comments online.
But then posts shared on a local Facebook page, followed by thousands of people, took a more sinister turn.
She discovered videos on the page of her singing and dancing. It showed her face, and an attached recording featuring offensive lyrics about her body.
The mum-of-two said she then faced a battle to get posts featured on a fake Facebook account in her name taken down.
When she contacted its parent company Meta, she said she was told the posts did not breach community standards.
Salt said they were eventually removed, but she was not sure whether they were taken down by Facebook or the administrators of the page.
In response, Meta said its manipulated media policy allowed the removal of videos that are the product of AI or machine learning, including deep learning techniques.
It said under its existing policies, once it is made aware of content, audio, photos or videos, “whether a deepfake or not, will be removed from Facebook if they violate any of our other Community Standards including those governing nudity, graphic violence, voter suppression and hate speech”.