The AI video creation field is undergoing a revolutionary transformation as the "Xunguang" platform launched by the DAMO Academy of Alibaba has made a stunning debut at the WAIC. This is not just a technical demonstration, but an all-in-one AI video creation platform that is set to open for beta testing and available for creators to use. It integrates multiple production processes such as scriptwriting, storyboarding, and video material editing into one platform, making video creation simpler and more efficient than ever before.
Imagine being able to easily convert selected targets into separate layers and effortlessly embed them into different background videos with a simple operation. The application of AI technology makes complex post-production tasks like adjusting camera angles, removing or modifying targets a one-click process, greatly improving production efficiency and allowing creators to focus more on creativity.
The core advantage of the "Xunguang" platform lies in its powerful AI editing capabilities, which not only understand complex plots but also maintain the consistency of generated objects, as well as enabling fine-grained editing and control of videos. Whether it's character control, scene control, or camera movement control, Xunguang can edit at the semantic level, not just simple pixel editing.
Main Functions:
Scriptwriting Assistance: The platform assists users in scriptwriting and storyboarding.
AI Editing Support: Supports a variety of AI editing functions for generated and uploaded materials.
Camera Movement Control: Supports control of video camera movements, such as translation, push-in/pull-out, and rotation.
Target Addition/Removal/Modification: Users can add, remove, or modify target objects in the video.
Video Layer Editing: Supports editing based on video layers, such as generating transparent background videos to merge with other backgrounds.
Layer Decomposition: Supports decomposing specific objects in the video into independent transparent background video layers.
Style Transfer: Provides over 20 different style transfer functions.
Facial Precision Control: Supports precise control of faces during video editing.
Fine-grained Editing and Control: Emphasizes fine-grained editing and control, including the editing and modification of local targets such as bodies, faces, foregrounds, and backgrounds in videos.
In addition to the video layer editing capabilities, the Xunguang platform also allows users to generate transparent background videos through text input and seamlessly merge them into other background videos with one click. It also supports layer decomposition functionality, which can automatically track and decompose targets in videos to form independent transparent background video layers.
The launch of the Xunguang platform is an important attempt by the DAMO Academy in the AI video field, showcasing its technical strength in AI video generation and editing as well as its deep understanding of the combination of AI productivity and human imagination. Through Xunguang, the DAMO Academy aims to reshape the video creation workflow, allowing creators to more efficiently turn creativity into reality.
With the global surge in AIGC, the launch of the Xunguang platform is a significant step in liberating productivity and advancing the video creation industry. It not only provides creators with a powerful tool but also points the way for the development of the entire industry. As the Xunguang platform continues to be refined and optimized, we have every reason to believe it will become an important milestone in the video creation field.
For official website beta application address: https://top.aibase.com/tool/xunguangshipinchuangzuopingtai