人工智能领域的又一里程碑式进展，OpenAI公司宣布，其最新的文本-图像融合模型——GPT-4 Turbo with Vision，现已通过API向全球开发者全面开放。这款模型不仅继承了GPT-4 Turbo系列模型的128,000个token窗口大小，还拥有截至2023年12月的知识库，更重要的是，它首次引入了视觉理解能力。
在以往，开发者处理文本和图像信息时，往往需要调用不同的模型，这不仅增加了工作量，也延长了开发周期。而GPT-4 Turbo with Vision的推出，将文本和图像处理合二为一，极大地简化了开发流程，同时也为未来的应用场景开拓了新的可能性。
这款模型能够理解并处理图像中的信息，从而使得人工智能不仅能够理解文本内容，还能分析图片、视频等视觉信息，实现了从文本到视觉的全面覆盖。这对于需要结合视觉和文本信息的应用场景，如图像描述、视频理解、内容推荐等，具有重大意义。
OpenAI的这一举措，无疑将进一步推动人工智能技术的发展，为各行各业带来创新和变革。随着技术的不断进步，我们期待着人工智能在视觉理解方面的能力能够得到进一步的提升，从而更好地服务于人类社会。
英语如下：
Title: “AI Breakthrough: GPT-4 Turbo with Vision Fully Opened to API”
Keywords: Visual AI, GPT-4, Easy Development
News Content:
Another milestone in the field of artificial intelligence, OpenAI has announced that its latest text-image fusion model – GPT-4 Turbo with Vision – is now fully open to global developers through an API. This model not only inherits the 128,000-token window size of the GPT-4 Turbo series but also has a knowledge base up to December 2023. What’s more significant is its introduction of visual understanding capabilities.
In the past, developers had to call different models to handle text and image information, which not only increased workload but also lengthened the development cycle. The launch of GPT-4 Turbo with Vision integrates text and image processing, significantly simplifying the development process and opening up new possibilities for future applications.
This model can understand and process information in images, making artificial intelligence capable not only of understanding text content but also analyzing visual information such as pictures and videos, covering the full spectrum from text to visuals. This is of great significance for applications that require a combination of visual and textual information, such as image captioning, video understanding, and content recommendation.
OpenAI’s move is undoubtedly set to further drive the development of artificial intelligence technology, bringing innovation and change to all walks of life. As technology continues to advance, we look forward to seeing further improvements in the visual understanding capabilities of AI, which will better serve humanity.
【来源】https://www.ithome.com/0/760/898.htm
Views: 2