OpenAI推出专用AI检测工具 识别DALL-E生成图像准确率达98%
财联社报道，当地时间周二，美国科技公司OpenAI发布公告称，为了区别照片是由相机还是自家AI生成的，公司推出了一款图像检测分类器——用来分析一幅图像是由OpenAI的“文生图”工具DALL·E 3生成的概率。公司从周二开始接受研究人员访问该工具的申请。
据悉，DALL·E 3是OpenAI公司开发的一款基于深度学习的图像生成模型，可以根据文本描述生成相应的图像。然而，随着DALL·E 3技术的不断发展，如何区分一幅图像是由AI生成的还是由相机拍摄的成为了亟待解决的问题。为了解决这一问题，OpenAI推出了这款专用AI检测工具。
据OpenAI官方表示，这款图像检测分类器的准确率达到了98%，可以有效地识别出由DALL·E 3生成的图像。该工具的推出意味着研究人员可以在学术研究和实际应用中，更加方便地使用DALL·E 3技术，同时也为版权保护和真实性验证提供了有力支持。
此外，OpenAI还表示，他们将继续改进这款检测工具，提高其准确率和稳定性，以满足更多场景的需求。研究人员可以通过申请访问该工具，开展相关领域的创新研究，为社会发展做出贡献。
这款图像检测分类器的推出，展示了OpenAI在人工智能领域的深厚实力和创新能力。未来，随着人工智能技术的不断进步，相信OpenAI还会带来更多令人惊艳的产品和服务。
英语如下：
**Title:** OpenAI Unveils Image Detection Tool With 98% Accuracy
**Keywords:** OpenAI, Image Detection, DALL-E
**News Content:**
*OpenAI Launches Specialized AI Detection Tool; Identifies DALL-E Generated Images With 98% Accuracy*
Financial Express reported that on Tuesday, US tech company OpenAI announced that in order to distinguish between photos taken by cameras and those generated by its own AI, it had launched an image detection classifier. This tool is designed to analyze the probability that an image was generated by OpenAI’s “text-to-image” tool DALL·E 3. The company began accepting applications from researchers to access the tool on Tuesday.
It is understood that DALL·E 3 is an image generation model developed by OpenAI based on deep learning. It can generate corresponding images based on text descriptions. However, as the technology of DALL·E 3 continues to evolve, distinguishing between images generated by AI and those captured by cameras has become an urgent issue. To address this problem, OpenAI has launched this specialized AI detection tool.
According to OpenAI’s official statement, this image detection classifier has an accuracy rate of 98%, effectively identifying images generated by DALL·E 3. The launch of this tool means that researchers can more conveniently use DALL·E 3 technology in academic research and practical applications. It also provides strong support for copyright protection and authenticity verification.
In addition, OpenAI also stated that they will continue to improve this detection tool, increasing its accuracy and stability to meet the needs of more scenarios. Researchers can apply to access the tool to carry out innovative research in their respective fields, contributing to social development.
The launch of this image detection classifier showcases OpenAI’s profound strength and innovative capabilities in the field of artificial intelligence. In the future, with the continuous advancement of artificial intelligence technology, it is believed that OpenAI will bring more impressive products and services.
【来源】https://www.cls.cn/detail/1669768
Views: 2