"Debiasing Vision-Language Models via Biased Prompts. (arXiv:2302.00070v1 [cs.LG])" — A general approach for debiasing vision-language foundation models by projecting out biased directions in the text embedding by debiasing only the text embedding with a calibrated projection matrix to yield robust classifiers and fair generative models.

Paper: http://arxiv.org/abs/2302.00070
Code: https://github.com/chingyaoc/debias_vl

#AI #CV #NewPaper #DeepLearning #MachineLearning

<<Find this useful? Please boost so that others can benefit too 🙂>>
Bias of Generative Models. The …