May 13, 2011 - At the Shanghai stop of the FORCE LINK AI Innovation Tour held today, theVolcano Engineofficially released a series of AI model upgrades, includingBean curd・Video Generation Model Seedance 1.0 lite, Beanbag 1.5・Visual Deep Thinking Model, and upgraded Beanbag・Music Model, aiming to help enterprises open up the application chain from business to intelligent body through a more comprehensive model matrix and richer intelligent body tools.
According to 1AI, the release of Seedance 1.0 lite is a small parameter number version of the beanbag video generation model series, the model supports text-generated video and graph-generated video generation in two ways, the video generation duration can be selected from 5 or 10 seconds, and the resolution provides two options of 480P and 720P. Enterprise users can use the model's API interface on the Volcano Ark platform, while individual users can experience it through the Doubao App or the Dream platform.
According to the official statement, Seedance 1.0 lite has realized two major breakthroughs in technology:First, the quality of movie and TV-grade video generation, and second, the dramatic increase in video generation speedThe Seedance 1.0 lite is a new model of the Seedance 1.0 lite, which balances effect, speed and cost-effectiveness. The model has a more accurate command following ability. With powerful semantic understanding, Seedance 1.0 lite can finely control the details of the character's appearance, dress style, expression and action, and has advantages in multi-subject action parsing, embedded text response, degree adverbs, and camera switching response.
also,Seedance 1.0 lite also supports a wide range of movie and TV-grade camera technologies.The video generated by the model is characterized by fine HD basic picture quality and movie-level aesthetics, including 360-degree surround, aerial photography, zoom, panning, following, handheld, and other lens languages. In terms of motion interaction, the model improves the natural and smooth interaction between characters and objects by analyzing the timing and spatial relationship of actions, so that the motion trajectory and force feedback are more in line with the laws of reality.
Seedance 1.0 lite model has a wide range of applications, including e-commerce advertising, entertainment effects, movie and TV creation, dynamic wallpaper and other fields. For example, in the field of e-commerce, the model can help merchants quickly generate high-quality marketing video materials, accurately matching the product display and activity promotion scenarios, thus effectively reducing production costs and cycles.