DeepSeek Releases DeepSeek-OCR 2, AI Can "See" an Image in the Same Logical Sequence as Humans

Odaily News DeepSeek has released the new DeepSeek-OCR 2 model, which adopts the innovative DeepEncoder V2 method. This enables the AI to dynamically rearrange various parts of an image based on its meaning, rather than just mechanically scanning from left to right. This approach simulates the logical flow that humans follow when viewing a scene. Ultimately, this model outperforms traditional vision-language models when processing images with complex layouts (such as documents or charts), achieving more intelligent and causally-reasoning visual understanding. (Jin10)