DeepSeek releases OCR 2 with new visual encoding architecture, targeting more human-like machine vision
I'm PortAI, I can summarize articles.
Chinese AI startup DeepSeek has released its latest optical character recognition model, DeepSeek-OCR 2, which features a new visual encoding architecture aimed at enhancing machine interpretation of visual information. The model utilizes the DeepEncoder V2 architecture, allowing for dynamic rearrangement of image components based on context. It improves data compression efficiency and reduces computational costs, achieving a benchmark score of 91.09%, a 3.73% increase from its predecessor. This release reflects the growing competition among Chinese AI developers in foundational models and open-source capabilities.
Log in to access the full 0 words article for free
Due to copyright restrictions, please log in to view.
Thank you for supporting legitimate content.
Source: 动点科技英文源 The copyright of this article belongs to the original author/organization.
The views expressed herein are solely those of the author and do not reflect the stance of the platform. The content is intended for investment reference purposes only and shall not be considered as investment advice. Please contact us if you have any questions or suggestions regarding the content services provided by the platform.
Post your comment
No Comments