Tweeted By @jekbradbury

Some fun adversarial ML happening in this unbelievably comprehensive @citizenlab analysis of image censorship on WeChat https://t.co/qqt0Ej9BDd

— James Bradbury (@jekbradbury) August 16, 2018