San Francisco, July 31 Micro-blogging site Twitter has detailed a new bounty competition that offers prizes of up to $3,500 for showing biases in its automatic image crops.
The winning teams will receive cash prizes via HackerOne $3,500 for the winner, $1,000 for second place, $500 for third place $1,000 each for Most Innovative and Most Generalisable.
"In May, we shared our approach to identifying bias in our saliency algorithm (also known as our image cropping algorithm), and we made our code available for others to reproduce our work," Rumman Chowdhury, Director, Software Engineering at Twitter, said in a blog post on Friday.
"We want to take this work a step further by inviting and incentivising the community to help identify potential harms of this algorithm beyond what we identified ourselves," Chowdhury added.
According to Chowdhury, they are inspired by how the research and hacker communities helped the security field establish best practices for identifying and mitigating vulnerabilities to protect the public.
"We want to cultivate a similar community, focused on ML ethics, to help us identify a broader range of issues than we would be able to on our own," Chowdhury said.
"With this challenge, we aim to set a precedent at Twitter, and in the industry, for proactive and collective identification of algorithmic harms," Chowdhury added.
For this challenge, Twitter said they are re-sharing their saliency model and the code used to generate a crop of an image given a predicted maximally salient point and asking participants to build their assessment.
Successful entries will consider both quantitative and qualitative methods in their approach.
All participants must enroll with HackerOne to make a valid submission; anyone with a HackerOne account may participate in this challenge.
Thr winners will be announced at the DEF CON AI Village workshop hosted by Twitter on August 8, where Twitter will invite the winners to present their work.
Disclaimer: This post has been auto-published from an agency feed without any modifications to the text and has not been reviewed by an editor