Zhixiang 3.0 (智象)

Trained on SenseTime’s “SenseCore” (商汤大装置) AI super-computer “Zhixiang multi-modal generation large model version 3.0 has comprehensively upgraded image and video generation capabilities.” According to Mei Tao, founder and CEO of Zhixiang Future, it specifically includes improved picture quality and relevance, more controllable lens movement and picture movement, and multi-scene-driven optimization. At the same time, Zhixiang Future also released the Zhixiang Multimodal Understanding Large Model version 1.0, which provides a more accurate and detailed understanding of images and video content through image modeling at the object level and spatiotemporal modeling at the event level. In the latest iterative version, the Smart Image Creator Platform strengthens the natural language interaction with users and is committed to implementation “if you can type, you can make videos”. According to reports, on the basis of the original Wensheng video, the Zhixiang Creator Platform has added adjustments to voice command input in the video part, and at the same time, it can be organized into relevant model output commands based on the video content uploaded by the user. This feature will greatly lower the learning threshold for users to edit videos using AIGC tools.