DeImage
Coming Soon
DeImage is DeData's platform dedicated to image annotation and training data for large-scale image models. Through intelligent distribution, reward evaluation, and user profiling, the platform smartly assigns annotation tasks to Web3 annotators, achieving high-efficiency image labeling and training. DeImage is capable of annotating simple images and also preprocessing complex images, making the original data easier and more user-friendly for annotation.
Apply DeImage Cases
Users are abtle to apple a case on DeImage: users can choose language and role that they prefer
Language: English and Simplified Chinese
Role:
Maker: users need to summarize the Image, Provide a description for the Image
Checker: Users need to vote on all captions, choose the best captions. only the highest votes on captions will be the final news result
Generate captions
Maker users can use a paragraph to describe the picture based on the content of the picture. if the Captions get the highest votes, the Maker will get the highest points reward
Estimated Point: the points that user will receive if the Caption get the highest votes
Expire Time: Users need to finish the cases during the expiration time, if the cases expired, users need to apply cases again
Signature onChain
After users finish the captions, user need to sign on the wallet(Metamask, Trust wallet, Okx wallet etc.)
Checker voting
If user choose checker role, users need to vote on all the cases captions. If the first caption get the highest votes, all checker who vote on the first caption will get the highest points. Other checkers will get lower points or 0 point if doing evil Rewards
Makers and checkers will receive points after the cases was finished, which mean if the case is not finished, there will no pints shown on the dashboard. so no need to be so harry, just wait for the cases finished by all users, it usually take less than 5mins
Last updated
