"Science" journal introduces Proofig AI to detect image fraud and avoid academic fraud

IT House announced in a news release on January 5 that the journal "Science" is introducing a commercial AI software called Proofig, which claims to be able to automatically detect "fraudulent images" in papers. , to prevent academic fraud. This initiative is expected to improve the quality and credibility of academic research.

The pictures in the journal "Science" are from official press releases.

To give an example of academic research, there is a study called "Western blot" published in the journal "Science". This study used some images to support the experimental data, but the journal pointed out that the researchers' process of citing the images lacked evidence, which provided opportunities for people with ulterior motives to modify the content of the images, which could lead to "fraudulent academic research." It should be noted that the study was not accused of fraud.

IT Home learned from the press release that the "Science" journal team claimed to have used Proofig software for several months of testing. This AI software is said to be able to accurately identify images that have been cropped, rotated, and changed contrast, and from which Circle potentially "fraudulent" images for further review by human editors.

In addition, the "Science" journal team also plans to launch a weekly news report collection called "Policy Pak", which will cover a variety of highly timely research information.

The above is the detailed content of "Science" journal introduces Proofig AI to detect image fraud and avoid academic fraud. For more information, please follow other related articles on the PHP Chinese website!